AI For Business Strategy Mit

AI For Business Strategy Mit — independent reviews, comparisons, pricing and step-by-step guides on Aizhi.

  • Harris corner detector

    Harris corner detector

    The Harris corner detector is a corner detection operator that is commonly used in computer vision algorithms to extract corners and infer features of an image. It was first introduced by Chris Harris and Mike Stephens in 1988 upon the improvement of Moravec's corner detector. Compared to its predecessor, Harris' corner detector takes the differential of the corner score into account with reference to direction directly, instead of using shifting patches for every 45 degree angles, and has been proved to be more accurate in distinguishing between edges and corners. Since then, it has been improved and adopted in many algorithms to preprocess images for subsequent applications. == Introduction == A corner is a point whose local neighborhood stands in two dominant and different edge directions. In other words, a corner can be interpreted as the junction of two edges, where an edge is a sudden change in image brightness. Corners are the important features in the image, and they are generally termed as interest points which are invariant to translation, rotation and illumination. Although corners are only a small percentage of the image, they contain the most important features in restoring image information, and they can be used to minimize the amount of processed data for motion tracking, image stitching, building 2D mosaics, stereo vision, image representation and other related computer vision areas. In order to capture the corners from the image, researchers have proposed many different corner detectors including the Kanade-Lucas-Tomasi (KLT) operator and the Harris operator which are most simple, efficient and reliable for use in corner detection. These two popular methodologies are both closely associated with and based on the local structure matrix. Compared to the Kanade-Lucas-Tomasi corner detector, the Harris corner detector provides good repeatability under changing illumination and rotation, and therefore, it is more often used in stereo matching and image database retrieval. Although there still exist drawbacks and limitations, the Harris corner detector is still an important and fundamental technique for many computer vision applications. == Development of Harris corner detection algorithm == Source: Without loss of generality, we will assume a grayscale 2-dimensional image is used. Let this image be given by I {\displaystyle I} . Consider taking an image patch ( x , y ) ∈ W {\displaystyle (x,y)\in W} (window) and shifting it by ( Δ x , Δ y ) {\displaystyle (\Delta x,\Delta y)} . The sum of squared differences (SSD) between these two patches, denoted f {\displaystyle f} , is given by: f ( Δ x , Δ y ) = ∑ ( x k , y k ) ∈ W ( I ( x k , y k ) − I ( x k + Δ x , y k + Δ y ) ) 2 {\displaystyle f(\Delta x,\Delta y)={\underset {(x_{k},y_{k})\in W}{\sum }}\left(I(x_{k},y_{k})-I(x_{k}+\Delta x,y_{k}+\Delta y)\right)^{2}} I ( x + Δ x , y + Δ y ) {\displaystyle I(x+\Delta x,y+\Delta y)} can be approximated by a Taylor expansion. Let I x {\displaystyle I_{x}} and I y {\displaystyle I_{y}} be the partial derivatives of I {\displaystyle I} , such that I ( x + Δ x , y + Δ y ) ≈ I ( x , y ) + I x ( x , y ) Δ x + I y ( x , y ) Δ y {\displaystyle I(x+\Delta x,y+\Delta y)\approx I(x,y)+I_{x}(x,y)\Delta x+I_{y}(x,y)\Delta y} This produces the approximation f ( Δ x , Δ y ) ≈ ∑ ( x , y ) ∈ W ( I x ( x , y ) Δ x + I y ( x , y ) Δ y ) 2 , {\displaystyle f(\Delta x,\Delta y)\approx {\underset {(x,y)\in W}{\sum }}\left(I_{x}(x,y)\Delta x+I_{y}(x,y)\Delta y\right)^{2},} which can be written in matrix form: f ( Δ x , Δ y ) ≈ ( Δ x Δ y ) M ( Δ x Δ y ) , {\displaystyle f(\Delta x,\Delta y)\approx {\begin{pmatrix}\Delta x&\Delta y\end{pmatrix}}M{\begin{pmatrix}\Delta x\\\Delta y\end{pmatrix}},} where M is the structure tensor, M = ∑ ( x , y ) ∈ W [ I x 2 I x I y I x I y I y 2 ] = [ ∑ ( x , y ) ∈ W I x 2 ∑ ( x , y ) ∈ W I x I y ∑ ( x , y ) ∈ W I x I y ∑ ( x , y ) ∈ W I y 2 ] {\displaystyle M={\underset {(x,y)\in W}{\sum }}{\begin{bmatrix}I_{x}^{2}&I_{x}I_{y}\\I_{x}I_{y}&I_{y}^{2}\end{bmatrix}}={\begin{bmatrix}{\underset {(x,y)\in W}{\sum }}I_{x}^{2}&{\underset {(x,y)\in W}{\sum }}I_{x}I_{y}\\{\underset {(x,y)\in W}{\sum }}I_{x}I_{y}&{\underset {(x,y)\in W}{\sum }}I_{y}^{2}\end{bmatrix}}} == Process of Harris corner detection algorithm == Commonly, Harris corner detector algorithm can be divided into five steps. Color to grayscale Spatial derivative calculation Structure tensor setup Harris response calculation Non-maximum suppression === Color to grayscale === If we use Harris corner detector in a color image, the first step is to convert it into a grayscale image, which will enhance the processing speed. The value of the gray scale pixel can be computed as a weighted sums of the values R, B and G of the color image, ∑ C ∈ { R , G , B } w C ⋅ C {\displaystyle \sum _{C\,\in \,\{R,G,B\}}w_{C}\cdot C} , where, e.g., w R = 0.299 , w G = 0.587 , w B = 1 − ( w R + w G ) = 0.114. {\displaystyle w_{R}=0.299,\ w_{G}=0.587,\ w_{B}=1-(w_{R}+w_{G})=0.114.} === Spatial derivative calculation === Next, we are going to find the derivative with respect to x and the derivative with respect to y, I x ( x , y ) {\displaystyle I_{x}(x,y)} and I y ( x , y ) {\displaystyle I_{y}(x,y)} . This can be approximated by applying Sobel operators. === Structure tensor setup === With I x ( x , y ) {\displaystyle I_{x}(x,y)} , I y ( x , y ) {\displaystyle I_{y}(x,y)} , we can construct the structure tensor M {\displaystyle M} . === Harris response calculation === For x ≪ y {\displaystyle x\ll y} , one has x ⋅ y x + y = x 1 1 + x / y ≈ x . {\displaystyle {\tfrac {x\cdot y}{x+y}}=x{\tfrac {1}{1+x/y}}\approx x.} In this step, we compute the smallest eigenvalue of the structure tensor using that approximation: λ min ≈ λ 1 λ 2 ( λ 1 + λ 2 ) = det ( M ) tr ⁡ ( M ) {\displaystyle \lambda _{\min }\approx {\frac {\lambda _{1}\lambda _{2}}{(\lambda _{1}+\lambda _{2})}}={\frac {\det(M)}{\operatorname {tr} (M)}}} with the trace t r ( M ) = m 11 + m 22 {\displaystyle \mathrm {tr} (M)=m_{11}+m_{22}} . Another commonly used Harris response calculation is shown as below, R = λ 1 λ 2 − k ( λ 1 + λ 2 ) 2 = det ( M ) − k tr ⁡ ( M ) 2 {\displaystyle R=\lambda _{1}\lambda _{2}-k(\lambda _{1}+\lambda _{2})^{2}=\det(M)-k\operatorname {tr} (M)^{2}} where k {\displaystyle k} is an empirically determined constant; k ∈ [ 0.04 , 0.06 ] {\displaystyle k\in [0.04,0.06]} . === Non-maximum suppression === In order to pick up the optimal values to indicate corners, we find the local maxima as corners within the window which is a 3 by 3 filter. == Improvement == Sources: Harris-Laplace Corner Detector Differential Morphological Decomposition Based Corner Detector Multi-scale Bilateral Structure Tensor Based Corner Detector == Applications == Image Alignment, Stitching and Registration 2D Mosaics Creation 3D Scene Modeling and Reconstruction Motion Detection Object Recognition Image Indexing and Content-based Retrieval Video Tracking

    Read more →
  • Language engineering

    Language engineering

    Language engineering involves the creation of natural language processing systems, whose cost and outputs are measurable and predictable. It is a distinct field contrasted to natural language processing and computational linguistics. A recent trend of language engineering is the use of Semantic Web technologies for the creation, archiving, processing, and retrieval of machine processable language data. Meta-Language Engineering is a proposed extension of Language Engineering first recorded in 2025, associated with the work of Delyone de Paula Canedo Filho. The term is used to designate an approach that, in addition to natural language processing, encompasses the symbolic, cognitive, and epistemological structuring of language systems.

    Read more →
  • Deep Instinct

    Deep Instinct

    Deep Instinct is a cybersecurity company that applies deep learning to cybersecurity. The company implements artificial intelligence to the task of preventing and detecting malware. The company was the recipient of the Technology Pioneer by The World Economic Forum in 2017. Lane Bess has been CEO of the company since 2022. == Overview == In 2015, Deep Instinct was founded by Guy Caspi, Dr. Eli David, and Nadav Maman. The headquarters of the company is located in New York City. In July 2017, NVIDIA became an investor. According to Tom's Hardware, NVIDIA’s investment enabled access to a GPU-based neural network and CUDA platform, which they were using to achieve maximum vulnerability detection rates. As of February 2020, the company had raised $43 million in Series C funding round. In April 2021, Deep Instinct raised $100 million in Series D funding to accelerate growth. == Partnerships == In April 2019, Deep Instinct partnered with Chinese artist, Guo O. Dong on an art project titled, The Persistence of Chaos, consisting of a laptop infected with 6 pieces of malware that represented $95 billion in damages. The art was auctioned with a final bid of $1,345,000. In the same year, Globes reported that, HP Inc partnered with Deep Instinct to launch their security solution HP SureSense, which has been applied to the EliteBook and Zbook devices.

    Read more →
  • Randomized Hough transform

    Randomized Hough transform

    Hough transforms are techniques for object detection, a critical step in many implementations of computer vision, or data mining from images. Specifically, the Randomized Hough transform is a probabilistic variant to the classical Hough transform, and is commonly used to detect curves (straight line, circle, ellipse, etc.) The basic idea of Hough transform (HT) is to implement a voting procedure for all potential curves in the image, and at the termination of the algorithm, curves that do exist in the image will have relatively high voting scores. Randomized Hough transform (RHT) is different from HT in that it tries to avoid conducting the computationally expensive voting process for every nonzero pixel in the image by taking advantage of the geometric properties of analytical curves, and thus improve the time efficiency and reduce the storage requirement of the original algorithm. == Motivation == Although Hough transform (HT) has been widely used in curve detection, it has two major drawbacks: First, for each nonzero pixel in the image, the parameters for the existing curve and redundant ones are both accumulated during the voting procedure. Second, the accumulator array (or Hough space) is predefined in a heuristic way. The more accuracy needed, the higher parameter resolution should be defined. These two needs usually result in a large storage requirement and low speed for real applications. Therefore, RHT was brought up to tackle this problem. == Implementation == In comparison with HT, RHT takes advantage of the fact that some analytical curves can be fully determined by a certain number of points on the curve. For example, a straight line can be determined by two points, and an ellipse (or a circle) can be determined by three points. The case of ellipse detection can be used to illustrate the basic idea of RHT. The whole process generally consists of three steps: Fit ellipses with randomly selected points. Update the accumulator array and corresponding scores. Output the ellipses with scores higher than some predefined threshold. === Ellipse fitting === One general equation for defining ellipses is: a ( x − p ) 2 + 2 b ( x − p ) ( y − q ) + c ( y − q ) 2 = 1 {\displaystyle a(x-p)^{2}+2b(x-p)(y-q)+c(y-q)^{2}=1} with restriction: a c − b 2 > 0 {\displaystyle ac-b^{2}>0} However, an ellipse can be fully determined if one knows three points on it and the tangents in these points. RHT starts by randomly selecting three points on the ellipse. Let them be X 1 {\displaystyle X_{1}} , X 2 {\displaystyle X_{2}} and X 3 {\displaystyle X_{3}} . The first step is to find the tangents of these three points. They can be found by fitting a straight line using least squares technique for a small window of neighboring pixels. The next step is to find the intersection points of the tangent lines. This can be easily done by solving the line equations found in the previous step. Then let the intersection points be T 12 {\displaystyle T_{12}} and T 23 {\displaystyle T_{23}} , the midpoints of line segments X 1 X 2 {\displaystyle X_{1}X_{2}} and X 2 X 3 {\displaystyle X_{2}X_{3}} be M 12 {\displaystyle M_{12}} and M 23 {\displaystyle M_{23}} . Then the center of the ellipse will lie in the intersection of T 12 M 12 {\displaystyle T_{12}M_{12}} and T 23 M 23 {\displaystyle T_{23}M_{23}} . Again, the coordinates of the intersected point can be determined by solving line equations and the detailed process is skipped here for conciseness. Let the coordinates of ellipse center found in previous step be ( x 0 , y 0 ) {\displaystyle (x_{0},y_{0})} . Then the center can be translated to the origin with x ′ = x − x 0 {\displaystyle x'=x-x_{0}} and y ′ = y − y 0 {\displaystyle y'=y-y_{0}} so that the ellipse equation can be simplified to: a x ′ 2 + 2 b x ′ y ′ + c y ′ 2 = 1 {\displaystyle ax'^{2}+2bx'y'+cy'^{2}=1} Now we can solve for the rest of ellipse parameters: a {\displaystyle a} , b {\displaystyle b} and c {\displaystyle c} by substituting the coordinates of X 1 {\displaystyle X_{1}} , X 2 {\displaystyle X_{2}} and X 3 {\displaystyle X_{3}} into the equation above. === Accumulating === With the ellipse parameters determined from previous stage, the accumulator array can be updated correspondingly. Different from classical Hough transform, RHT does not keep "grid of buckets" as the accumulator array. Rather, it first calculates the similarities between the newly detected ellipse and the ones already stored in accumulator array. Different metrics can be used to calculate the similarity. As long as the similarity exceeds some predefined threshold, replace the one in the accumulator with the average of both ellipses and add 1 to its score. Otherwise, initialize this ellipse to an empty position in the accumulator and assign a score of 1. === Termination === Once the score of one candidate ellipse exceeds the threshold, it is determined as existing in the image (in other words, this ellipse is detected), and should be removed from the image and accumulator array so that the algorithm can detect other potential ellipses faster. The algorithm terminates when the number of iterations reaches a maximum limit or all the ellipses have been detected. Pseudo code for RHT: while (we find ellipses AND not reached the maximum epoch) { for (a fixed number of iterations) { Find a potential ellipse. if (the ellipse is similar to an ellipse in the accumulator) then Replace the one in the accumulator with the average of two ellipses and add 1 to the score; else Insert the ellipse into an empty position in the accumulator with a score of 1; } Select the ellipse with the best score and save it in a best ellipse table; Eliminate the pixels of the best ellipse from the image; Empty the accumulator; }

    Read more →
  • Cost-sensitive machine learning

    Cost-sensitive machine learning

    Cost-sensitive machine learning is an approach within machine learning that considers varying costs associated with different types of errors. This method diverges from traditional approaches by introducing a cost matrix, explicitly specifying the penalties or benefits for each type of prediction error. The inherent difficulty which cost-sensitive machine learning tackles is that minimizing different kinds of classification errors is a multi-objective optimization problem. == Overview == Cost-sensitive machine learning optimizes models based on the specific consequences of misclassifications, making it a valuable tool in various applications. It is especially useful in problems with a high imbalance in class distribution and a high imbalance in associated costs Cost-sensitive machine learning introduces a scalar cost function in order to find one (of multiple) Pareto optimal points in this multi-objective optimization problem (similar to the Weighted sum model) == Cost Matrix == The cost matrix is a crucial element within cost-sensitive modeling, explicitly defining the costs or benefits associated with different prediction errors in classification tasks. Represented as a table, the matrix aligns true and predicted classes, assigning a cost value to each combination. For instance, in binary classification, it may distinguish costs for false positives and false negatives. The utility of the cost matrix lies in its application to calculate the expected cost or loss. The formula, expressed as a double summation, utilizes joint probabilities: Expected Loss = ∑ i ∑ j P ( Actual i , Predicted j ) ⋅ Cost Actual i , Predicted j {\displaystyle {\text{Expected Loss}}=\sum _{i}\sum _{j}P({\text{Actual}}_{i},{\text{Predicted}}_{j})\cdot {\text{Cost}}_{{\text{Actual}}_{i},{\text{Predicted}}_{j}}} Here, P ( Actual i , Predicted j ) {\displaystyle P({\text{Actual}}_{i},{\text{Predicted}}_{j})} denotes the joint probability of actual class i {\displaystyle i} and predicted class j {\displaystyle j} , providing a nuanced measure that considers both the probabilities and associated costs. This approach allows practitioners to fine-tune models based on the specific consequences of misclassifications, adapting to scenarios where the impact of prediction errors varies across classes. == Applications == === Fraud Detection === In the realm of data science, particularly in finance, cost-sensitive machine learning is applied to fraud detection. By assigning different costs to false positives and false negatives, models can be fine-tuned to minimize the overall financial impact of misclassifications. === Medical Diagnostics === In healthcare, cost-sensitive machine learning plays a role in medical diagnostics. The approach allows for customization of models based on the potential harm associated with misdiagnoses, ensuring a more patient-centric application of machine learning algorithms. == Challenges == A typical challenge in cost-sensitive machine learning is the reliable determination of the cost matrix which may evolve over time. == Literature == Cost-Sensitive Machine Learning. USA, CRC Press, 2011. ISBN 9781439839287 Abhishek, K., Abdelaziz, D. M. (2023). Machine Learning for Imbalanced Data: Tackle Imbalanced Datasets Using Machine Learning and Deep Learning Techniques. (n.p.): Packt Publishing. ISBN 9781801070881

    Read more →
  • Superquadrics

    Superquadrics

    In mathematics, the superquadrics or super-quadrics (also superquadratics) are a family of geometric shapes defined by formulas that resemble those of ellipsoids and other quadrics, except that the squaring operations are replaced by arbitrary powers. They can be seen as the three-dimensional relatives of the superellipses. The term may refer to the solid object or to its surface, depending on the context. The equations below specify the surface; the solid is specified by replacing the equality signs by less-than-or-equal signs. The superquadrics include many shapes that resemble cubes, octahedra, cylinders, lozenges and spindles, with rounded or sharp corners. Because of their flexibility and relative simplicity, they are popular geometric modeling tools, especially in computer graphics. It becomes an important geometric primitive widely used in computer vision, robotics, and physical simulation. Some authors, such as Alan Barr, define "superquadrics" as including both the superellipsoids and the supertoroids. In modern computer vision literatures, superquadrics and superellipsoids are used interchangeably, since superellipsoids are the most representative and widely utilized shape among all the superquadrics. Comprehensive coverage of geometrical properties of superquadrics and methods of their recovery from range images and point clouds are covered in several computer vision literatures. == Formulas == === Implicit equation === The surface of the basic superquadric is given by | x | r + | y | s + | z | t = 1 {\displaystyle \left|x\right|^{r}+\left|y\right|^{s}+\left|z\right|^{t}=1} where r, s, and t are positive real numbers that determine the main features of the superquadric. Namely: less than 1: a pointy octahedron modified to have concave faces and sharp edges. exactly 1: a regular octahedron. between 1 and 2: an octahedron modified to have convex faces, blunt edges and blunt corners. exactly 2: a sphere greater than 2: a cube modified to have rounded edges and corners. infinite (in the limit): a cube Each exponent can be varied independently to obtain combined shapes. For example, if r=s=2, and t=4, one obtains a solid of revolution which resembles an ellipsoid with round cross-section but flattened ends. This formula is a special case of the superellipsoid's formula if (and only if) r = s. If any exponent is allowed to be negative, the shape extends to infinity. Such shapes are sometimes called super-hyperboloids. The basic shape above spans from -1 to +1 along each coordinate axis. The general superquadric is the result of scaling this basic shape by different amounts A, B, C along each axis. Its general equation is | x A | r + | y B | s + | z C | t = 1. {\displaystyle \left|{\frac {x}{A}}\right|^{r}+\left|{\frac {y}{B}}\right|^{s}+\left|{\frac {z}{C}}\right|^{t}=1.} === Parametric description === Parametric equations in terms of surface parameters u and v (equivalent to longitude and latitude if m equals 2) are x ( u , v ) = A g ( v , 2 r ) g ( u , 2 r ) y ( u , v ) = B g ( v , 2 s ) f ( u , 2 s ) z ( u , v ) = C f ( v , 2 t ) − π 2 ≤ v ≤ π 2 , − π ≤ u < π , {\displaystyle {\begin{aligned}x(u,v)&{}=Ag\left(v,{\frac {2}{r}}\right)g\left(u,{\frac {2}{r}}\right)\\y(u,v)&{}=Bg\left(v,{\frac {2}{s}}\right)f\left(u,{\frac {2}{s}}\right)\\z(u,v)&{}=Cf\left(v,{\frac {2}{t}}\right)\\&-{\frac {\pi }{2}}\leq v\leq {\frac {\pi }{2}},\quad -\pi \leq u<\pi ,\end{aligned}}} where the auxiliary functions are f ( ω , m ) = sgn ⁡ ( sin ⁡ ω ) | sin ⁡ ω | m g ( ω , m ) = sgn ⁡ ( cos ⁡ ω ) | cos ⁡ ω | m {\displaystyle {\begin{aligned}f(\omega ,m)&{}=\operatorname {sgn}(\sin \omega )\left|\sin \omega \right|^{m}\\g(\omega ,m)&{}=\operatorname {sgn}(\cos \omega )\left|\cos \omega \right|^{m}\end{aligned}}} and the sign function sgn(x) is sgn ⁡ ( x ) = { − 1 , x < 0 0 , x = 0 + 1 , x > 0. {\displaystyle \operatorname {sgn}(x)={\begin{cases}-1,&x<0\\0,&x=0\\+1,&x>0.\end{cases}}} === Spherical product === Barr introduces the spherical product which given two plane curves produces a 3D surface. If f ( μ ) = ( f 1 ( μ ) f 2 ( μ ) ) , g ( ν ) = ( g 1 ( ν ) g 2 ( ν ) ) {\displaystyle f(\mu )={\begin{pmatrix}f_{1}(\mu )\\f_{2}(\mu )\end{pmatrix}},\quad g(\nu )={\begin{pmatrix}g_{1}(\nu )\\g_{2}(\nu )\end{pmatrix}}} are two plane curves then the spherical product is h ( μ , ν ) = f ( μ ) ⊗ g ( ν ) = ( f 1 ( μ ) g 1 ( ν ) f 1 ( μ ) g 2 ( ν ) f 2 ( μ ) ) {\displaystyle h(\mu ,\nu )=f(\mu )\otimes g(\nu )={\begin{pmatrix}f_{1}(\mu )\ g_{1}(\nu )\\f_{1}(\mu )\ g_{2}(\nu )\\f_{2}(\mu )\end{pmatrix}}} This is similar to the typical parametric equation of a sphere: x = x 0 + r sin ⁡ θ cos ⁡ φ y = y 0 + r sin ⁡ θ sin ⁡ φ ( 0 ≤ θ ≤ π , 0 ≤ φ < 2 π ) z = z 0 + r cos ⁡ θ {\displaystyle {\begin{aligned}x&=x_{0}+r\sin \theta \;\cos \varphi \\y&=y_{0}+r\sin \theta \;\sin \varphi \qquad (0\leq \theta \leq \pi ,\;0\leq \varphi <2\pi )\\z&=z_{0}+r\cos \theta \end{aligned}}} which give rise to the name spherical product. Barr uses the spherical product to define quadric surfaces, like ellipsoids, and hyperboloids as well as the torus, superellipsoid, superquadric hyperboloids of one and two sheets, and supertoroids. == Plotting code == The following GNU Octave code generates a mesh approximation of a superquadric:

    Read more →
  • Concordancer

    Concordancer

    A concordancer is a computer program that automatically constructs a concordance—an alphabetised index of every occurrence of a word or phrase in a body of text, each entry displayed with its surrounding context. Concordancers are primary tools in corpus linguistics, lexicography, computer-assisted translation, and language teaching. The most common display format is the key word in context (KWIC) layout, in which each hit appears centred on a line with a fixed span of words to its left and right, enabling rapid scanning of usage patterns across many occurrences. == History == === Pre-computational concordances === The compilation of concordances predates computers by many centuries. Around 1230, the French Dominican cardinal Hugh of Saint-Cher directed a team of friars in assembling a concordance of the Latin Vulgate Bible, generally regarded as the first systematic concordance of any text. To help readers locate passages, Hugh divided each biblical chapter into lettered sections. Later milestones include a Hebrew Old Testament concordance compiled by Rabbi Mordecai Nathan (1448), Alexander Cruden's Complete Concordance to the Holy Scriptures (1737), and the manuscript Asaf ha-Mazkir, an unfinished concordance to the Babylonian Talmud compiled by Moses Rigotz around the turn of the 19th century. === First computer concordance === The first concordance produced with computing assistance was the Index Thomisticus, a comprehensive lexical index of the writings of and around Thomas Aquinas, totalling approximately 10.6 million Latin words. The Italian Jesuit priest Roberto Busa conceived the project in 1946 and secured the sponsorship of IBM in 1949 after a meeting with chairman Thomas J. Watson. Keypunch operators in Gallarate, Italy, encoded the texts onto punched cards from around 1950. IBM executive Paul Tasman developed the processing methods. The full 56-volume printed edition was completed around 1980, followed by a CD-ROM edition in 1989 and a web-accessible version in 2005. === The KWIC format === The key word in context (KWIC) display was formalised as a computational technique by Hans Peter Luhn, a researcher at IBM, in a 1960 paper in American Documentation. In KWIC output, each instance of the search term (the node word) is centred on a line with a fixed window of words to each side; sorting the resulting lines alphabetically by the immediately adjacent word reveals collocational and phraseological patterns at a glance. === COCOA === One of the first dedicated concordancing programs was COCOA (COunt and COncordance Generation on Atlas), created in 1965 by D. B. Russell at University College London and the Atlas Computer Laboratory in Harwell, Oxfordshire. Written in approximately 4,000 cards of FORTRAN, it processed text annotated with flat, non-hierarchical markup tags and could produce word counts and concordances in multiple languages. Within its first six months COCOA had been applied to texts in at least six languages. A second version designed for multiple mainframe platforms was distributed to British computing centres in the mid-1970s. Growing dissatisfaction with its interface and the eventual withdrawal of Atlas Laboratory support prompted British funding bodies to commission a successor program. === Oxford Concordance Program === The Oxford Concordance Program (OCP) was designed and written in FORTRAN by Susan Hockey and Ian Marriott at Oxford University Computing Services (OUCS) between 1979 and 1980 and first released in 1981. Hockey and Marriott acknowledged that OCP owed much to COCOA and the CLOC system at the University of Birmingham. OCP accepted COCOA-format markup to encode metadata such as author, act, scene, and line number, and was described by its authors as "a machine-independent text analysis program for producing word lists, indices and concordances in a variety of languages and alphabets." By the mid-1980s it had been licensed to approximately 240 institutions in 23 countries. A personal computer version, Micro-OCP, was developed for the IBM PC and sold by Oxford University Press from the late 1980s. Version 2 was rewritten in 1985–86 and documented in the same 1987 article by Hockey and co-author John Martin. === Personal computer era === The availability of affordable personal computers in the 1980s and 1990s enabled standalone concordancing applications that analysts could run locally without specialist computing facilities. MicroConcord, developed by Mike Scott and Tim Johns and published by Oxford University Press in 1993 for MS-DOS, was among the first concordancers designed specifically for classroom language teaching. WordSmith Tools, also developed by Mike Scott, was first released in 1996 and became one of the most widely used corpus analysis suites in academic linguistics research. Other tools from this era include TACT (University of Toronto, 1989), a suite of MS-DOS freeware programs for literary text analysis, and MonoConc, a Windows concordancer created by Michael Barlow. === Web-based concordancers === From the late 1990s onwards, web-based concordancers hosted on remote servers gave researchers browser access to large preloaded corpora without requiring local storage or processing. The Sketch Engine, developed by Adam Kilgarriff and Pavel Rychlý (Masaryk University), was launched commercially in July 2003 by Lexical Computing Limited and introduced word sketches—automatically generated one-page profiles of a word's typical grammatical relations and collocations. AntConc, created by Laurence Anthony at Waseda University, Tokyo, was first released in 2002 as freeware for Windows, macOS, and Linux. == Features == Modern concordancers typically offer a range of analytical functions beyond basic KWIC display. These commonly include: KWIC display with the node word centred and context words in aligned columns, sortable by the word one, two, or three positions to the left or right of the node (L1–L3 and R1–R3) Concordance plots, visualising the distribution of hits as marks along a scaled bar representing each text in the corpus Frequency and word lists, both alphabetical and ranked by frequency Collocation statistics, identifying words that co-occur with the search term more often than chance, quantified by measures such as mutual information, the t-score, or log-likelihood Keyword analysis, comparing word frequencies between a study corpus and a reference corpus to identify statistically distinctive items N-gram analysis, finding frequently recurring word sequences of a specified length Part-of-speech tagging integration, allowing searches filtered to particular grammatical categories Unicode support for multilingual text Bilingual and parallel concordancers additionally display aligned text in two or more languages side by side, enabling comparison of translation equivalents across language pairs. == Notable concordancers == === WordSmith Tools === Created by Mike Scott and first released in 1996, WordSmith Tools is a Windows corpus analysis suite that evolved from MicroConcord. Its three core modules are Concord (KWIC concordances), WordList (frequency and alphabetical word lists), and Keywords (statistical keyword identification relative to a reference corpus). Oxford University Press used WordSmith Tools for dictionary preparation work. Version 4.0 is freely available; later versions are sold by Lexical Analysis Software Limited. === AntConc === AntConc is a freeware, multiplatform concordancing toolkit created by Laurence Anthony, Professor of Applied Linguistics at Waseda University, Tokyo. First released in 2002 and formally described in a 2005 academic paper, it runs on Windows, macOS, and Linux. Its tools include a KWIC concordancer, a concordance plot for visualising distribution across texts, a collocates tool, a keyword list, and an n-gram analysis module. Because it is free and requires only plain text files, AntConc is widely used in linguistics courses and independent research worldwide. === Sketch Engine === The Sketch Engine is a corpus management and query system co-created by Adam Kilgarriff and Pavel Rychlý and launched in 2003 by Lexical Computing Limited. It provides browser-based access to over 800 corpora in more than 100 languages. Beyond concordance searching, it offers word sketches, collocation analysis, distributional thesaurus construction, keyword and terminology extraction, and diachronic analysis. It is used by major publishers including Macmillan and Oxford University Press for lexicographic research. A subset tool, SKELL (Sketch Engine for Language Learning), is freely accessible to individual learners. === Wmatrix === Wmatrix is a web-based corpus processing environment developed by Paul Rayson at the University Centre for Computer Corpus Research on Language (UCREL), Lancaster University. Alongside concordances and frequency lists, Wmatrix integrates CLAWS part-of-speech tagging and the USAS semantic tagger, enabling keyword analysis simultane

    Read more →
  • Superellipsoid

    Superellipsoid

    In mathematics, a superellipsoid (or super-ellipsoid) is a solid whose horizontal sections are superellipses (Lamé curves) with the same squareness parameter ϵ 2 {\displaystyle \epsilon _{2}} , and whose vertical sections through the center are superellipses with the squareness parameter ϵ 1 {\displaystyle \epsilon _{1}} . It is a generalization of an ellipsoid, which is a special case when ϵ 1 = ϵ 2 = 1 {\displaystyle \epsilon _{1}=\epsilon _{2}=1} . Superellipsoids as computer graphics primitives were popularized by Alan H. Barr (who used the name "superquadrics" to refer to both superellipsoids and supertoroids). In modern computer vision and robotics literatures, superquadrics and superellipsoids are used interchangeably, since superellipsoids are the most representative and widely utilized shape among all the superquadrics. Superellipsoids have a rich shape vocabulary, including cuboids, cylinders, ellipsoids, octahedra and their intermediates. It becomes an important geometric primitive widely used in computer vision, robotics, and physical simulation. The main advantage of describing objects and environment with superellipsoids is its conciseness and expressiveness in shape. Furthermore, a closed-form expression of the Minkowski sum between two superellipsoids is available. This makes it a desirable geometric primitive for robot grasping, collision detection, and motion planning. == Special cases == A handful of notable mathematical figures can arise as special cases of superellipsoids given the correct set of values, which are depicted in the above graphic: Cylinder Sphere Steinmetz solid Bicone Regular octahedron Cube, as a limiting case where the exponents tend to infinity Piet Hein's supereggs are also special cases of superellipsoids. == Formulas == === Basic (normalized) superellipsoid === The basic superellipsoid is defined by the implicit function f ( x , y , z ) = ( x 2 ϵ 2 + y 2 ϵ 2 ) ϵ 2 / ϵ 1 + z 2 ϵ 1 {\displaystyle f(x,y,z)=\left(x^{\frac {2}{\epsilon _{2}}}+y^{\frac {2}{\epsilon _{2}}}\right)^{\epsilon _{2}/\epsilon _{1}}+z^{\frac {2}{\epsilon _{1}}}} The parameters ϵ 1 {\displaystyle \epsilon _{1}} and ϵ 2 {\displaystyle \epsilon _{2}} are positive real numbers that control the squareness of the shape. The surface of the superellipsoid is defined by the equation: f ( x , y , z ) = 1 {\displaystyle f(x,y,z)=1} For any given point ( x , y , z ) ∈ R 3 {\displaystyle (x,y,z)\in \mathbb {R} ^{3}} , the point lies inside the superellipsoid if f ( x , y , z ) < 1 {\displaystyle f(x,y,z)<1} , and outside if f ( x , y , z ) > 1 {\displaystyle f(x,y,z)>1} . Any "parallel of latitude" of the superellipsoid (a horizontal section at any constant z between -1 and +1) is a Lamé curve with exponent 2 / ϵ 2 {\displaystyle 2/\epsilon _{2}} , scaled by a = ( 1 − z 2 ϵ 1 ) ϵ 1 2 {\displaystyle a=(1-z^{\frac {2}{\epsilon _{1}}})^{\frac {\epsilon _{1}}{2}}} , which is ( x a ) 2 ϵ 2 + ( y a ) 2 ϵ 2 = 1. {\displaystyle \left({\frac {x}{a}}\right)^{\frac {2}{\epsilon _{2}}}+\left({\frac {y}{a}}\right)^{\frac {2}{\epsilon _{2}}}=1.} Any "meridian of longitude" (a section by any vertical plane through the origin) is a Lamé curve with exponent 2 / ϵ 1 {\displaystyle 2/\epsilon _{1}} , stretched horizontally by a factor w that depends on the sectioning plane. Namely, if x = u cos ⁡ θ {\displaystyle x=u\cos \theta } and y = u sin ⁡ θ {\displaystyle y=u\sin \theta } , for a given θ {\displaystyle \theta } , then the section is ( u w ) 2 ϵ 1 + z 2 ϵ 1 = 1 , {\displaystyle \left({\frac {u}{w}}\right)^{\frac {2}{\epsilon _{1}}}+z^{\frac {2}{\epsilon _{1}}}=1,} where w = ( cos 2 ϵ 2 ⁡ θ + sin 2 ϵ 2 ⁡ θ ) − ϵ 2 2 . {\displaystyle w=(\cos ^{\frac {2}{\epsilon _{2}}}\theta +\sin ^{\frac {2}{\epsilon _{2}}}\theta )^{-{\frac {\epsilon _{2}}{2}}}.} In particular, if ϵ 2 {\displaystyle \epsilon _{2}} is 1, the horizontal cross-sections are circles, and the horizontal stretching w {\displaystyle w} of the vertical sections is 1 for all planes. In that case, the superellipsoid is a solid of revolution, obtained by rotating the Lamé curve with exponent 2 / ϵ 1 {\displaystyle 2/\epsilon _{1}} around the vertical axis. === Superellipsoid === The basic shape above extends from −1 to +1 along each coordinate axis. The general superellipsoid is obtained by scaling the basic shape along each axis by factors a x {\displaystyle a_{x}} , a y {\displaystyle a_{y}} , a z {\displaystyle a_{z}} , the semi-diameters of the resulting solid. The implicit function is F ( x , y , z ) = ( ( x a x ) 2 ϵ 2 + ( y a y ) 2 ϵ 2 ) ϵ 2 ϵ 1 + ( z a z ) 2 ϵ 1 {\displaystyle F(x,y,z)=\left(\left({\frac {x}{a_{x}}}\right)^{\frac {2}{\epsilon _{2}}}+\left({\frac {y}{a_{y}}}\right)^{\frac {2}{\epsilon _{2}}}\right)^{\frac {\epsilon _{2}}{\epsilon _{1}}}+\left({\frac {z}{a_{z}}}\right)^{\frac {2}{\epsilon _{1}}}} . Similarly, the surface of the superellipsoid is defined by the equation F ( x , y , z ) = 1 {\displaystyle F(x,y,z)=1} For any given point ( x , y , z ) ∈ R 3 {\displaystyle (x,y,z)\in \mathbb {R} ^{3}} , the point lies inside the superellipsoid if f ( x , y , z ) < 1 {\displaystyle f(x,y,z)<1} , and outside if f ( x , y , z ) > 1 {\displaystyle f(x,y,z)>1} . Therefore, the implicit function is also called the inside-outside function of the superellipsoid. The superellipsoid has a parametric representation in terms of surface parameters η ∈ [ − π / 2 , π / 2 ) {\displaystyle \eta \in [-\pi /2,\pi /2)} , ω ∈ [ − π , π ) {\displaystyle \omega \in [-\pi ,\pi )} . x ( η , ω ) = a x cos ϵ 1 ⁡ η cos ϵ 2 ⁡ ω {\displaystyle x(\eta ,\omega )=a_{x}\cos ^{\epsilon _{1}}\eta \cos ^{\epsilon _{2}}\omega } y ( η , ω ) = a y cos ϵ 1 ⁡ η sin ϵ 2 ⁡ ω {\displaystyle y(\eta ,\omega )=a_{y}\cos ^{\epsilon _{1}}\eta \sin ^{\epsilon _{2}}\omega } z ( η , ω ) = a z sin ϵ 1 ⁡ η {\displaystyle z(\eta ,\omega )=a_{z}\sin ^{\epsilon _{1}}\eta } === General posed superellipsoid === In computer vision and robotic applications, a superellipsoid with a general pose in the 3D Euclidean space is usually of more interest. For a given Euclidean transformation of the superellipsoid frame g = [ R ∈ S O ( 3 ) , t ∈ R 3 ] ∈ S E ( 3 ) {\displaystyle g=[\mathbf {R} \in SO(3),\mathbf {t} \in \mathbb {R} ^{3}]\in SE(3)} relative to the world frame, the implicit function of a general posed superellipsoid surface defined the world frame is F ( g − 1 ∘ ( x , y , z ) ) = 1 {\displaystyle F\left(g^{-1}\circ (x,y,z)\right)=1} where ∘ {\displaystyle \circ } is the transformation operation that maps the point ( x , y , z ) ∈ R 3 {\displaystyle (x,y,z)\in \mathbb {R} ^{3}} in the world frame into the canonical superellipsoid frame. === Volume of superellipsoid === The volume encompassed by the superelllipsoid surface can be expressed in terms of the beta functions β ( ⋅ , ⋅ ) {\displaystyle \beta (\cdot ,\cdot )} , V ( ϵ 1 , ϵ 2 , a x , a y , a z ) = 2 a x a y a z ϵ 1 ϵ 2 β ( ϵ 1 2 , ϵ 1 + 1 ) β ( ϵ 2 2 , ϵ 2 + 2 2 ) {\displaystyle V(\epsilon _{1},\epsilon _{2},a_{x},a_{y},a_{z})=2a_{x}a_{y}a_{z}\epsilon _{1}\epsilon _{2}\beta ({\frac {\epsilon _{1}}{2}},\epsilon _{1}+1)\beta ({\frac {\epsilon _{2}}{2}},{\frac {\epsilon _{2}+2}{2}})} or equivalently with the Gamma function Γ ( ⋅ ) {\displaystyle \Gamma (\cdot )} , since β ( m , n ) = Γ ( m ) Γ ( n ) Γ ( m + n ) {\displaystyle \beta (m,n)={\frac {\Gamma (m)\Gamma (n)}{\Gamma (m+n)}}} == Recovery from data == Recoverying the superellipsoid (or superquadrics) representation from raw data (e.g., point cloud, mesh, images, and voxels) is an important task in computer vision, robotics, and physical simulation. Traditional computational methods model the problem as a least-square problem. The goal is to find out the optimal set of superellipsoid parameters θ ≐ [ ϵ 1 , ϵ 2 , a x , a y , a z , g ] {\displaystyle \theta \doteq [\epsilon _{1},\epsilon _{2},a_{x},a_{y},a_{z},g]} that minimize an objective function. Other than the shape parameters, g ∈ {\displaystyle g\in } SE(3) is the pose of the superellipsoid frame with respect to the world coordinate. There are two commonly used objective functions. The first one is constructed directly based on the implicit function G 1 ( θ ) = a x a y a z ∑ i = 1 N ( F ϵ 1 ( g − 1 ∘ ( x i , y i , z i ) ) − 1 ) 2 {\displaystyle G_{1}(\theta )=a_{x}a_{y}a_{z}\sum _{i=1}^{N}\left(F^{\epsilon _{1}}\left(g^{-1}\circ (x_{i},y_{i},z_{i})\right)-1\right)^{2}} The minimization of the objective function provides a recovered superellipsoid as close as possible to all the input points { ( x i , y i , z i ) ∈ R 3 , i = 1 , 2 , . . . , N } {\displaystyle \{(x_{i},y_{i},z_{i})\in \mathbb {R} ^{3},i=1,2,...,N\}} . At the mean time, the scalar value a x , a y , a z {\displaystyle a_{x},a_{y},a_{z}} is positively proportional to the volume of the superellipsoid, and thus have the effect of minimizing the volume as well. The other objective function tries to minimized the radial distance between the points and the superellipsoid. That is G 2 ( θ ) = ∑ i = 1 N ( | r

    Read more →
  • Automatic acquisition of lexicon

    Automatic acquisition of lexicon

    Automatic acquisition of lexicon is a computerized process used for the development of a complex morphological lexicon of a language. The lexicon is essential for the NLP (Natural language processing), as well as a prerequisite to any wide-coverage parser. The two main requirements represent raw corpus and the morphological description of the language. The aim is to provide lemmas that will serve to the explanation of all the words that occur within the corpus. For the achievement of a quality lexicon it is necessary to manually validate the generated lemmas and iterate the whole process several times. The process is focused on the open word classes (e.g. nouns, adjectives, verbs). Closed classes (e.g. prepositions, pronouns, numerals) are excluded. This method is applicable to the languages with a rich morphology, such as Slovak, Russian or Croatian. Applied to Slovak, being an inflectional language, the automatic acquisition focuses on the inflectional morphology as well as on the derivational morphology. This fact enables the users to find out the information about derivational relations (e.g. adjectivizations, prefixes) in the lexicon. For example, Slovak word korpusový is an adjectivization of korpus (eng. corpus). == Three-step loop == Conformably to Benoît Sagot, there are three stages involved in the acquisition of lemmas: Generation and inflection Ranking Manual validation The more iteration will be performed, the more accurate lexicon will be obtained. For each iteration are essential the information given by a manual validator. === Generation and inflection === Firstly, all words which represent the closed word classes (pronouns, prepositions, numerals) are manually excluded from the given corpus. Number of their occurrences in the corpus is provided. Then the automatic generation comes, when the hypothetical lemmas according to the morphological description of a language are created. Generated lemmas are consequently being inflected, so that all of their inflected forms are built. Obtained forms are associated with the corresponding lemma and a morphological tag. === Ranking === There was created a probabilistic model, represented by a fix-point algorithm, to rank the hypothetical lemmas generated in the first step. Best ranked lemmas are expected to be ideally all correct, whereas the least ranked tend to be incorrect. === Manual validation === Correctness of the best- ranked lemmas created in the previous step are checked by the manual validator, who should be a native speaker. Lemmas are at this stage divided into three categories: valid lemmas, appended to lexicon erroneous lemmas generated by valid forms (later associated to another lemmas) erroneous lemmas generated by invalid forms (these need to be excluded) == Future development == Automatic acquisition, in comparison to a purely manual development of the lexicons, seems to be promising, considering the future development, because of the short validation time needed and the relatively small amount of human labor involved.

    Read more →
  • Egocentric vision

    Egocentric vision

    Egocentric vision or first-person vision is a sub-field of computer vision that entails analyzing images and videos captured by a wearable camera, which is typically worn on the head or on the chest and naturally approximates the visual field of the camera wearer. Consequently, visual data capture the part of the scene on which the user focuses to carry out the task at hand and offer a valuable perspective to understand the user's activities and their context in a naturalistic setting. The wearable camera looking forwards is often supplemented with a camera looking inward at the user's eye and able to measure a user's eye gaze, which is useful to reveal attention and to better understand the user's activity and intentions. == History == The idea of using a wearable camera to gather visual data from a first-person perspective dates back to the 70s, when Steve Mann invented "Digital Eye Glass", a device that, when worn, causes the human eye itself to effectively become both an electronic camera and a television display. Subsequently, wearable cameras were used for health-related applications in the context of Humanistic Intelligence and Wearable AI. Egocentric vision is best done from the point-of-eye, but may also be done by way of a neck-worn camera when eyeglasses would be in-the-way. This neck-worn variant was popularized by way of the Microsoft SenseCam in 2006 for experimental health research works. The interest of the computer vision community into the egocentric paradigm has been arising slowly entering the 2010s and it is rapidly growing in recent years, boosted by both the impressive advances in the field of wearable technology and by the increasing number of potential applications. The prototypical first-person vision system described by Kanade and Hebert, in 2012 is composed by three basic components: a localization component able to estimate the surrounding, a recognition component able to identify object and people, and an activity recognition component, able to provide information about the current activity of the user. Together, these three components provide a complete situational awareness of the user, which in turn can be used to provide assistance to the user or to the caregiver. Following this idea, the first computational techniques for egocentric analysis focused on hand-related activity recognition and social interaction analysis. Also, given the unconstrained nature of the video and the huge amount of data generated, temporal segmentation and summarization were among the first problems addressed. After almost ten years of egocentric vision (2007–2017), the field is still undergoing diversification. Emerging research topics include: Social saliency estimation Multi-agent egocentric vision systems Privacy preserving techniques and applications Attention-based activity analysis Social interaction analysis Hand pose analysis Ego graphical User Interfaces (EUI) Understanding social dynamics and attention Revisiting robotic vision and machine vision as egocentric sensing Activity forecasting Gaze prediction == Technical challenges == Today's wearable cameras are small and lightweight digital recording devices that can acquire images and videos automatically, without the user intervention, with different resolutions and frame rates, and from a first-person point of view. Therefore, wearable cameras are naturally primed to gather visual information from our everyday interactions since they offer an intimate perspective of the visual field of the camera wearer. Depending on the frame rate, it is common to distinguish between photo-cameras (also called lifelogging cameras) and video-cameras. The former (e.g., Narrative Clip and Microsoft SenseCam), are commonly worn on the chest, and are characterized by a very low frame rate (up to 2fpm) that allows to capture images over a long period of time without the need of recharging the battery. Consequently, they offer considerable potential for inferring knowledge about e.g. behaviour patterns, habits or lifestyle of the user. However, due to the low frame-rate and the free motion of the camera, temporally adjacent images typically present abrupt appearance changes so that motion features cannot be reliably estimated. The latter (e.g., Google Glass, GoPro), are commonly mounted on the head, and capture conventional video (around 35fps) that allows to capture fine temporal details of interactions. Consequently, they offer potential for in-depth analysis of daily or special activities. However, since the camera is moving with the wearer head, it becomes more difficult to estimate the global motion of the wearer and in the case of abrupt movements, the images can result blurred. In both cases, since the camera is worn in a naturalistic setting, visual data present a huge variability in terms of illumination conditions and object appearance. Moreover, the camera wearer is not visible in the image and what he/she is doing has to be inferred from the information in the visual field of the camera, implying that important information about the wearer, such for instance as pose or facial expression estimation, is not available. == Applications == A collection of studies published in a special theme issue of the American Journal of Preventive Medicine has demonstrated the potential of lifelogs captured through wearable cameras from a number of viewpoints. In particular, it has been shown that used as a tool for understanding and tracking lifestyle behaviour, lifelogs would enable the prevention of noncommunicable diseases associated to unhealthy trends and risky profiles (such as obesity and depression). In addition, used as a tool of re-memory cognitive training, lifelogs would enable the prevention of cognitive and functional decline in elderly people. More recently, egocentric cameras have been used to study human and animal cognition, human-human social interaction, human-robot interaction, human expertise in complex tasks. Other applications include navigation/assistive technologies for the blind, monitoring and assistance of industrial workflows, and augmented reality interfaces.

    Read more →
  • 1.58-bit large language model

    1.58-bit large language model

    A 1.58-bit large language model (also known as a ternary LLM) is a type of large language model (LLM) designed to be computationally efficient. It achieves this by using weights that are restricted to only three values: -1, 0, and +1. This restriction significantly reduces the model's memory footprint and allows for faster processing, as computationally expensive multiplication operations can be replaced with lower-cost additions. This contrasts with traditional models that use 16-bit floating-point numbers (FP16 or BF16) for their weights. Studies have shown that for models up to several billion parameters, the performance of 1.58-bit LLMs on various tasks is comparable to their full-precision counterparts. This approach could enable powerful AI to run on less specialized and lower-power hardware. The name "1.58-bit" comes from the fact that a system with three states contains log 2 ⁡ 3 ≈ 1.58 {\displaystyle \log _{2}3\approx 1.58} bits of information. These models are sometimes also referred to as 1-bit LLMs in research papers, although this term can also refer to true binary models (with weights of -1 and +1). == BitNet == In 2024, Ma et al., researchers at Microsoft, declared that their 1.58-bit model, BitNet b1.58 is comparable in performance to the 16-bit Llama 2 and opens the era of 1-bit LLM. BitNet creators did not use the post-training quantization of weights but instead relied on the new BitLinear transform that replaced the nn.Linear layer of the traditional transformer design. In 2025, Microsoft researchers had released an open-weights and open inference code model BitNet b1.58 2B4T demonstrating performance competitive with the full precision models at 2B parameters and 4T training tokens. == Post-training quantization == BitNet derives its performance from being trained natively in 1.58 bit instead of being quantized from a full-precision model after training. Still, training is an expensive process and it would be desirable to be able to somehow convert an existing model to 1.58 bits. In 2024, HuggingFace reported a way to gradually ramp up the 1.58-bit quantization in fine-tuning an existing model down to 1.58 bits. == Critique == Some researchers point out that the scaling laws of large language models favor the low-bit weights only in case of undertrained models. As the number of training tokens increases, the deficiencies of low-bit quantization surface.

    Read more →
  • Brill tagger

    Brill tagger

    The Brill tagger is an inductive method for part-of-speech tagging. It was described and invented by Eric Brill in his 1993 PhD thesis. It can be summarized as an "error-driven transformation-based tagger". It is: a form of supervised learning, which aims to minimize error; and, a transformation-based process, in the sense that a tag is assigned to each word and changed using a set of predefined rules. In the transformation process, if the word is known, it first assigns the most frequent tag, or if the word is unknown, it naively assigns the tag "noun" to it. High accuracy is eventually achieved by applying these rules iteratively and changing the incorrect tags. This approach ensures that valuable information such as the morphosyntactic construction of words is employed in an automatic tagging process. == Algorithm == The algorithm starts with initialization, which is the assignment of tags based on their probability for each word (for example, "dog" is more often a noun than a verb). Then "patches" are determined via rules that correct (probable) tagging errors made in the initialization phase: Initialization: Known words (in vocabulary): assigning the most frequent tag associated to a form of the word Unknown word == Rules and processing == The input text is first tokenized, or broken into words. Typically in natural language processing, contractions such as "'s", "n't", and the like are considered separate word tokens, as are punctuation marks. A dictionary and some morphological rules then provide an initial tag for each word token. For example, a simple lookup would reveal that "dog" may be a noun or a verb (the most frequent tag is simply chosen), while an unknown word will be assigned some tag(s) based on capitalization, various prefix or suffix strings, etc. (such morphological analyses, which Brill calls Lexical Rules, may vary between implementations). After all word tokens have (provisional) tags, contextual rules apply iteratively, to correct the tags by examining small amounts of context. This is where the Brill method differs from other part of speech tagging methods such as those using Hidden Markov Models. Rules are reapplied repeatedly, until a threshold is reached, or no more rules can apply. Brill rules are of the general form: tag1 → tag2 IF Condition where the Condition tests the preceding and/or following word tokens, or their tags (the notation for such rules differs between implementations). For example, in Brill's notation: IN NN WDPREVTAG DT while would change the tag of a word from IN (preposition) to NN (common noun), if the preceding word's tag is DT (determiner) and the word itself is "while". This covers cases like "all the while" or "in a while", where "while" should be tagged as a noun rather than its more common use as a conjunction (many rules are more general). Rules should only operate if the tag being changed is also known to be permissible, for the word in question or in principle (for example, most adjectives in English can also be used as nouns). Rules of this kind can be implemented by simple Finite-state machines. See Part of speech tagging for more general information including descriptions of the Penn Treebank and other sets of tags. Typical Brill taggers use a few hundred rules, which may be developed by linguistic intuition or by machine learning on a pre-tagged corpus. == Code == Brill's code pages at Johns Hopkins University are no longer on the web. An archived version of a mirror of the Brill tagger at its latest version as it was available at Plymouth Tech can be found on Archive.org. The software uses the MIT License.

    Read more →
  • Linux color management

    Linux color management

    Linux color management has the same goal as the color management systems (CMS) for other operating systems, which is to achieve the best possible color reproduction throughout an imaging workflow from its source (camera, video, scanner, etc.), through imaging software (Digikam, darktable, RawTherapee, GIMP, Krita, Scribus, etc.), and finally onto an output medium (monitor, video projector, printer, etc.). In particular, color management attempts to enable color consistency across media and throughout a color-managed workflow. Linux color management relies on the use of accurate ICC (International Color Consortium) and DCP (DNG Color Profile) profiles describing the behavior of input and output devices, and color-managed applications that are aware of these profiles. These applications perform gamut conversions between device profiles and color spaces. Gamut conversions, based on accurate device profiles, are the essence of color management. Historically, color management was not an initial design consideration of the X Window System on which much of Linux graphics support rests, and thus color-managed workflows have been somewhat more challenging to implement on Linux than on other OS's such as Microsoft Windows or macOS. This situation is now being progressively remedied, and color management under Linux, while functional, has not yet acquired mature status. Although it is now possible to obtain a consistent color-managed workflow under Linux, certain problems still remain: The absence of a central user control panel for color settings. Some hardware devices for color calibration lack Linux drivers, firmware or accessory data. Since ICC color profiles are written to an open specification, they are compatible across operating systems. Hence, a profile produced on one OS should work on any other OS given the availability of the necessary software to read it and perform the gamut conversions. This can be used as a workaround for the lack of support for certain spectrophotometers or colorimeters under Linux: one can simply produce a profile on a different OS and then use it in a Linux workflow. Additionally, certain hardware, such as most printers and certain monitors, can be calibrated under another OS and then used in a fully color-managed workflow on Linux. The popular Ubuntu Linux distribution added initial color management in the 11.10 release (the "Oneiric Ocelot" release). == Requirements for a color-managed workflow == Accurate device profiles obtained with source or output characterization software. Correctly loaded video card lookup tables (LUTs) (or monitor profiles that do not require LUT adjustments). Color-managed applications that are configured to use a correct monitor profile and input/output profiles, with support for control over the rendering intent and black point compensation. Calibration and profiling requires: for input devices (scanner, camera, etc.) a color target which the profiling software will compare to the manufacturer-provided color values of the target. or for output devices (monitor, printer, etc.) a reading with a specific device (spectrophotometer, colorimeter or spectrocolorimeter) of the color patch values and comparing the measured values against the values originally sent for output. === Monitor calibration and profiling === One of the critical elements in any color-managed workflow is the monitor, because, at one step or another, handling and making color adaptation through imaging software is required for most images, thus the ability of the monitor to present accurate colors is crucial. Monitor color management consists of calibration and profiling. The first step, calibration, is done by adjusting the monitor controls and the output of the graphics card (via calibration curves) to match user-definable characteristics, such as brightness, white point and gamma. The calibration settings are stored in a .cal file. The second step, profiling (characterization), involves measuring the calibrated display's response and recording it in a color profile. The profile is stored in an .icc file ("ICC file"). For convenience, the calibration settings are usually stored together with the profile in the ICC file. Note that .icm files are identical to .icc files - the difference is only in the name. Seeing correct colors requires using a monitor profile-aware application, together with the same calibration used when profiling the monitor. Calibration alone does not yield accurate colors. If a monitor was calibrated before it was profiled, the profile will only yield correct colors when used on the monitor with the same calibration (the same monitor control adjustments and the same calibration curves loaded into the video card's lookup table). macOS has built-in support for loading calibration curves and installing a system-wide color profile. Windows 7 onward allows loading calibration curves, though this functionality must be enabled manually. Linux and older versions of Windows require using a standalone LUT loader. === Device profiles === ICC profiles are cross-platform and can thus be created on other operating systems and used under Linux. Monitor profiles, however, require some additional attention. Since a monitor profile depends both on the monitor itself and on the video card, a monitor profile should only be used with the same monitor and video card with which it was created. The monitor settings should not be adjusted after creating the profile. In addition, since most calibration software use LUT adjustments during calibration, the corresponding LUTs must be loaded every time the display server (X11, Wayland) is started (e.g. with every graphical login). In the unlikely case of a colorimeter being unsupported by Linux, a profile created under Windows or macOS can be used under Linux. === Display-channel lookup tables === There are two approaches to loading display channel LUTs: Create a profile that does not modify video card LUTs and thus does not require LUTs be loaded later on. Ideally, this approach would rely on DDC-capable monitors—the internal monitor settings of which are set via calibration software. Unfortunately, monitors capable of making these adjustments through DDC are not common and are generally expensive. There is only one calibration software on Linux that can interact with a DDC monitor. For mainstream monitors, a couple of options exist: BasICColor software, which works with most colorimeters on the market, allows one to adjust display output via the monitor interface, and then to choose a "Profile, do not calibrate" option. By doing this, one can create a profile that does not require video card LUT adjustments. For EyeOne devices, EyeOne Match allows the user to calibrate to "Native" gamma and white point targets, which results in the LUT adjustment curves displayed after the calibration as a simple, linear 1:1 mapping (a straight line from corner to corner). Both BasICColor and EyeOne Match do not presently run under Linux but they are capable of creating a profile that does not require LUT adjustments. Use an LUT loader to actually load the LUT adjustments contained within the profile prepared during calibration. According to the documentation, these loaders do not modify the video card LUT by itself, but achieve the same type of adjustment by modifying the X server gamma ramp. Loaders are available for Linux distributions that use X.org or XFree86—the two most popular X servers on Linux. Other X servers are not guaranteed to work with the currently available loaders. There are two LUT loaders available for Linux: Xcalib is one such loader, and although it is a command-line utility, it is quite easy to use. dispwin is a part of Argyll CMS. If, for any reason, the LUT cannot be loaded, it is still recommended to go through the initial stages of calibration where a user is asked by calibration software to make some manual adjustments to the monitor, as this will often improve display linearity and also provide information on its color temperature. This is especially recommended for CRT monitors. === Color-managed applications === In ICC-aware applications, it is important to make sure the correct profiles are assigned to devices, mainly to the monitor and the printer. Some Linux applications can auto-detect the monitor profile, while others requires that it is specified manually. Although there is no designated place to store device profiles on Linux, /usr/share/color/icc/ has become the de facto standard. Most applications running under WINE have not been fully tested for color accuracy. While 8-bpp programs can have some color resolution difficulties due to depth conversion errors, colors in higher-depth applications should be accurate, as long as those programs perform their gamut conversions based on the same monitor profile as that used for loading the LUT, granted that the corresponding LUT adjustments are loaded. == List of color-managed applications == darktabl

    Read more →
  • Lose It!

    Lose It!

    Lose It! is an American health and wellness mobile app developed by FitNow, Inc. The app generates calorie budgets for users by tracking weight, exercise, food and calorie intake, and personal goals, primarily to assist them in achieving weight loss. == History == Lose It! was developed in Boston and debuted in 2008. The app and its associated company were founded by J.J. Allaire, Charles Teague and Paul Dicristina. Prior to founding Lose It!, Teague and Allaire had founded the online research tool Onfolio, which was acquired by Microsoft in 2006. The Lose It! app was originally released as an iOS app before being released as a website in 2010 and an Android app in 2011. In 2015, Lose It! announced plans to release the app internationally. Lose It! was also available as an app for Apple Watch at its launch in 2015. The app’s “Snap It” feature, which allows users to approximate calorie counts by taking pictures of their daily meals and snacks, was released in beta in 2016. Snap It was named an Innovation Awards Honoree at the 2017 Consumer Electronics Show in Las Vegas. In 2020, Patrick Wetherille, one of the company’s earliest employees, was appointed chief executive officer. == App == Lose It! is weight loss app. The app allows users to set goals such as increasing strength, overall health/maintenance, and weight loss. It provides users recommended calorie budgets based on data such as their current weight and their desired weight. Lose It! also tracks data such as exercise/activity level and food consumption and allows users to track calories consumed by scanning barcodes for food products then retrieving calorie information for products. The app can also estimate the amount of calories in a food products. Lose It! has integration features connecting it to other apps such as Fitbit and Runkeeper. It also has social features such as joining groups and sharing progress with friends. The Premium version of the app allows users to track foods according to specific diets like keto, heart healthy or Mediterranean.

    Read more →
  • Neural style transfer

    Neural style transfer

    Neural style transfer (NST) software algorithms are able to manipulate digital images, or videos, in order to adopt the appearance or visual style of another image. NST algorithms are characterized by their use of deep neural networks for the sake of image transformation. Common uses for NST are the creation of artificial artwork from photographs, for example by transferring the appearance of famous paintings to user-supplied photographs. Several notable mobile apps use NST techniques for this purpose, including DeepArt and Prisma. This method has been used by artists and designers around the globe to develop new artwork based on existent style(s). == History == NST is an example of image stylization, a problem studied for over two decades within the field of non-photorealistic rendering. The first two example-based style transfer algorithms were image analogies and image quilting. Both of these methods were based on patch-based texture synthesis algorithms. Given a training pair of images–a photo and an artwork depicting that photo–a transformation could be learned and then applied to create new artwork from a new photo, by analogy. If no training photo was available, it would need to be produced by processing the input artwork; image quilting did not require this processing step, though it was demonstrated on only one style. NST was first published in the paper "A Neural Algorithm of Artistic Style" by Leon Gatys et al., originally released to ArXiv 2015, and subsequently accepted by the peer-reviewed CVPR conference in 2016. The original paper used a VGG-19 architecture that has been pre-trained to perform object recognition using the ImageNet dataset. In 2017, Google AI introduced a method that allows a single deep convolutional style transfer network to learn multiple styles at the same time. This algorithm permits style interpolation in real-time, even when done on video media. == Mathematics == This section closely follows the original paper. === Overview === The idea of Neural Style Transfer (NST) is to take two images—a content image p → {\displaystyle {\vec {p}}} and a style image a → {\displaystyle {\vec {a}}} —and generate a third image x → {\displaystyle {\vec {x}}} that minimizes a weighted combination of two loss functions: a content loss L content ( p → , x → ) {\displaystyle {\mathcal {L}}_{\text{content }}({\vec {p}},{\vec {x}})} and a style loss L style ( a → , x → ) {\displaystyle {\mathcal {L}}_{\text{style }}({\vec {a}},{\vec {x}})} . The total loss is a linear sum of the two: L NST ( p → , a → , x → ) = α L content ( p → , x → ) + β L style ( a → , x → ) {\displaystyle {\mathcal {L}}_{\text{NST}}({\vec {p}},{\vec {a}},{\vec {x}})=\alpha {\mathcal {L}}_{\text{content}}({\vec {p}},{\vec {x}})+\beta {\mathcal {L}}_{\text{style}}({\vec {a}},{\vec {x}})} By jointly minimizing the content and style losses, NST generates an image that blends the content of the content image with the style of the style image. Both the content loss and the style loss measures the similarity of two images. The content similarity is the weighted sum of squared-differences between the neural activations of a single convolutional neural network (CNN) on two images. The style similarity is the weighted sum of Gram matrices within each layer (see below for details). The original paper used a VGG-19 CNN, but the method works for any CNN. === Symbols === Let x → {\textstyle {\vec {x}}} be an image input to a CNN. Let F l ∈ R N l × M l {\textstyle F^{l}\in \mathbb {R} ^{N_{l}\times M_{l}}} be the matrix of filter responses in layer l {\textstyle l} to the image x → {\textstyle {\vec {x}}} , where: N l {\textstyle N_{l}} is the number of filters in layer l {\textstyle l} ; M l {\textstyle M_{l}} is the height times the width (i.e. number of pixels) of each filter in layer l {\textstyle l} ; F i j l ( x → ) {\textstyle F_{ij}^{l}({\vec {x}})} is the activation of the i th {\textstyle i^{\text{th}}} filter at position j {\textstyle j} in layer l {\textstyle l} . A given input image x → {\textstyle {\vec {x}}} is encoded in each layer of the CNN by the filter responses to that image, with higher layers encoding more global features, but losing details on local features. === Content loss === Let p → {\textstyle {\vec {p}}} be an original image. Let x → {\textstyle {\vec {x}}} be an image that is generated to match the content of p → {\textstyle {\vec {p}}} . Let P l {\textstyle P^{l}} be the matrix of filter responses in layer l {\textstyle l} to the image p → {\textstyle {\vec {p}}} . The content loss is defined as the squared-error loss between the feature representations of the generated image and the content image at a chosen layer l {\displaystyle l} of a CNN: L content ( p → , x → , l ) = 1 2 ∑ i , j ( A i j l ( x → ) − A i j l ( p → ) ) 2 {\displaystyle {\mathcal {L}}_{\text{content }}({\vec {p}},{\vec {x}},l)={\frac {1}{2}}\sum _{i,j}\left(A_{ij}^{l}({\vec {x}})-A_{ij}^{l}({\vec {p}})\right)^{2}} where A i j l ( x → ) {\displaystyle A_{ij}^{l}({\vec {x}})} and A i j l ( p → ) {\displaystyle A_{ij}^{l}({\vec {p}})} are the activations of the i th {\displaystyle i^{\text{th}}} filter at position j {\displaystyle j} in layer l {\displaystyle l} for the generated and content images, respectively. Minimizing this loss encourages the generated image to have similar content to the content image, as captured by the feature activations in the chosen layer. The total content loss is a linear sum of the content losses of each layer: L content ( p → , x → ) = ∑ l v l L content ( p → , x → , l ) {\displaystyle {\mathcal {L}}_{\text{content }}({\vec {p}},{\vec {x}})=\sum _{l}v_{l}{\mathcal {L}}_{\text{content }}({\vec {p}},{\vec {x}},l)} , where the v l {\displaystyle v_{l}} are positive real numbers chosen as hyperparameters. === Style loss === The style loss is based on the Gram matrices of the generated and style images, which capture the correlations between different filter responses at different layers of the CNN: L style ( a → , x → ) = ∑ l = 0 L w l E l , {\displaystyle {\mathcal {L}}_{\text{style }}({\vec {a}},{\vec {x}})=\sum _{l=0}^{L}w_{l}E_{l},} where E l = 1 4 N l 2 M l 2 ∑ i , j ( G i j l ( x → ) − G i j l ( a → ) ) 2 . {\displaystyle E_{l}={\frac {1}{4N_{l}^{2}M_{l}^{2}}}\sum _{i,j}\left(G_{ij}^{l}({\vec {x}})-G_{ij}^{l}({\vec {a}})\right)^{2}.} Here, G i j l ( x → ) {\displaystyle G_{ij}^{l}({\vec {x}})} and G i j l ( a → ) {\displaystyle G_{ij}^{l}({\vec {a}})} are the entries of the Gram matrices for the generated and style images at layer l {\displaystyle l} . Explicitly, G i j l ( x → ) = ∑ k F i k l ( x → ) F j k l ( x → ) {\displaystyle G_{ij}^{l}({\vec {x}})=\sum _{k}F_{ik}^{l}({\vec {x}})F_{jk}^{l}({\vec {x}})} Minimizing this loss encourages the generated image to have similar style characteristics to the style image, as captured by the correlations between feature responses in each layer. The idea is that activation pattern correlations between filters in a single layer captures the "style" on the order of the receptive fields at that layer. Similarly to the previous case, the w l {\displaystyle w_{l}} are positive real numbers chosen as hyperparameters. === Hyperparameters === In the original paper, they used a particular choice of hyperparameters. The style loss is computed by w l = 0.2 {\displaystyle w_{l}=0.2} for the outputs of layers conv1_1, conv2_1, conv3_1, conv4_1, conv5_1 in the VGG-19 network, and zero otherwise. The content loss is computed by w l = 1 {\displaystyle w_{l}=1} for conv4_2, and zero otherwise. The ratio α / β ∈ [ 5 , 50 ] × 10 − 4 {\displaystyle \alpha /\beta \in [5,50]\times 10^{-4}} . === Training === Image x → {\displaystyle {\vec {x}}} is initially approximated by adding a small amount of white noise to input image p → {\displaystyle {\vec {p}}} and feeding it through the CNN. Then we successively backpropagate this loss through the network with the CNN weights fixed in order to update the pixels of x → {\displaystyle {\vec {x}}} . After several thousand epochs of training, an x → {\displaystyle {\vec {x}}} (hopefully) emerges that matches the style of a → {\displaystyle {\vec {a}}} and the content of p → {\displaystyle {\vec {p}}} . As of 2017, when implemented on a GPU, it takes a few minutes to converge. == Extensions == In some practical implementations, it is noted that the resulting image has too much high-frequency artifact, which can be suppressed by adding the total variation to the total loss. Compared to VGGNet, AlexNet does not work well for neural style transfer. NST has also been extended to videos. Subsequent work improved the speed of NST for images by using special-purpose normalizations. In a paper by Fei-Fei Li et al. adopted a different regularized loss metric and accelerated method for training to produce results in real-time (three orders of magnitude faster than Gatys). Their idea was to use not the pixel-based loss defined above but rather a 'perceptual loss' measuring t

    Read more →