AI Chatbot Robot

AI Chatbot Robot — independent reviews, comparisons, pricing and step-by-step guides on Aizhi.

• 

  Object model

  In computing, object model has two related but distinct meanings: The properties of objects in general in a specific computer programming language, technology, notation or methodology that uses them. Examples are the object models of Java, the Component Object Model (COM), or Object-Modeling Technique (OMT). Such object models are usually defined using concepts such as class, generic function, message, inheritance, polymorphism, and encapsulation. There is an extensive literature on formalized object models as a subset of the formal semantics of programming languages. A collection of objects or classes through which a program can examine and manipulate some specific parts of its world. In other words, the object-oriented interface to some service or system. Such an interface is said to be the object model of the represented service or system. For example, the Document Object Model (DOM) is a collection of objects that represent a page in a web browser, used by script programs to examine and dynamically change the page. There is a Microsoft Excel object model [1] for controlling Microsoft Excel from another program, and the ASCOM Telescope Driver is an object model for controlling an astronomical telescope. == Features == An object model consists of the following important features: === Object reference === Objects can be accessed via object references. To invoke a method in an object, the object reference and method name are given, together with any arguments. === Interfaces === An interface provides a definition of the signature of a set of methods without specifying their implementation. An object will provide a particular interface if its class contains code that implement the method of that interface. An interface also defines types that can be used to declare the type of variables or parameters and return values of methods. === Actions === An action in object-oriented programming (OOP) is initiated by an object invoking a method in another object. An invocation can include additional information needed to carry out the method. The receiver executes the appropriate method and then returns control to the invoking object, sometimes supplying a result. === Exceptions === Programs can encounter various errors and unexpected conditions of varying seriousness. During the execution of the method many different problems may be discovered. Exceptions provide a clean way to deal with error conditions without complicating the code. A block of code may be defined to throw an exception whenever particular unexpected conditions or errors arise. This means that control passes to another block of code that catches the exception.

  Read more →
• 

  Dan Roth

  Dan Roth (Hebrew: דן רוט) is the Eduardo D. Glandt Distinguished Professor of Computer and Information Science at the University of Pennsylvania and the Chief AI Scientist at Oracle. Until June 2024 Roth was a VP and distinguished scientist at AWS AI. In his role at AWS, Roth led over the last three years the scientific effort behind the first-generation Generative AI products from AWS, including Titan Models, Amazon Q efforts, and Bedrock, from inception until they became generally available. Roth got his B.A. summa cum laude in mathematics from the Technion, Israel, and his Ph.D. in computer science from Harvard University in 1995. He taught at the University of Illinois at Urbana-Champaign from 1998 to 2017 before moving to the University of Pennsylvania. == Professional career == Roth is a Fellow of the American Association for the Advancement of Science (AAAS), the Association for Computing Machinery (ACM), the Association for the Advancement of Artificial Intelligence (AAAI), and the Association of Computational Linguistics (ACL). Roth’s research focuses on the computational foundations of intelligent behavior. He develops theories and systems pertaining to intelligent behavior using a unified methodology, at the heart of which is the idea that learning has a central role in intelligence. His work centers around the study of machine learning and inference methods to facilitate natural language understanding. In doing that he has pursued several interrelated lines of work that span multiple aspects of this problem - from fundamental questions in learning and inference and how they interact, to the study of a range of natural language processing (NLP) problems and developing advanced machine learning based tools for natural language applications. Roth has made seminal contribution to the fusion of Learning and Reasoning, Machine Learning with weak, incidental supervision, and to machine learning and inference approaches to natural language understanding. He has written the first paper on zero-shot learning in natural language processing, a 2008 paper by Chang, Ratinov, Roth, and Srikumar that was published at AAAI’08, but the name given to the learning paradigm there was dataless classification. Roth has worked on probabilistic reasoning (including its complexity and probabilistic lifted inference ), Constrained Conditional Models (ILP formulations of NLP problems) and constraints-driven learning, part-based (constellation) methods in object recognition, response based Learning, He has developed NLP and Information extraction tools that are being used broadly by researchers and commercially, including NER, coreference resolution, wikification, SRL, and ESL text correction. Roth is a co-founder of NexLP, Inc., a startup that applies natural language processing and machine learning in the legal and compliance domains. In 2020, NexLP was acquired by Reveal, Inc., an e-discovery software company. He is currently on the scientific advisory board of the Allen Institute for AI.

  Read more →
• 

  Statistical machine translation

  Statistical machine translation (SMT) is a machine translation approach where translations are generated on the basis of statistical models whose parameters are derived from the analysis of bilingual text corpora. The statistical approach contrasts with the rule-based approaches to machine translation as well as with example-based machine translation, that superseded the previous rule-based approach that required explicit description of each and every linguistic rule, which was costly, and which often did not generalize to other languages. The first ideas of statistical machine translation were introduced by Warren Weaver in 1949, including the ideas of applying Claude Shannon's information theory. Statistical machine translation was re-introduced in the late 1980s and early 1990s by researchers at IBM's Thomas J. Watson Research Center. Before the introduction of neural machine translation, it was by far the most widely studied machine translation method. == Basis == The idea behind statistical machine translation comes from information theory. A document is translated according to the probability distribution p ( e | f ) {\displaystyle p(e|f)} that a string e {\displaystyle e} in the target language (for example, English) is the translation of a string f {\displaystyle f} in the source language (for example, French). The problem of modeling the probability distribution p ( e | f ) {\displaystyle p(e|f)} has been approached in a number of ways. One approach which lends itself well to computer implementation is to apply Bayes' theorem, that is p ( e | f ) ∝ p ( f | e ) p ( e ) {\displaystyle p(e|f)\propto p(f|e)p(e)} , where the translation model p ( f | e ) {\displaystyle p(f|e)} is the probability that the source string is the translation of the target string, and the language model p ( e ) {\displaystyle p(e)} is the probability of seeing that target language string. This decomposition is attractive as it splits the problem into two subproblems. Finding the best translation e ~ {\displaystyle {\tilde {e}}} is done by picking up the one that gives the highest probability: e ~ = a r g max e ∈ e ∗ p ( e | f ) = a r g max e ∈ e ∗ p ( f | e ) p ( e ) {\displaystyle {\tilde {e}}=arg\max _{e\in e^{}}p(e|f)=arg\max _{e\in e^{}}p(f|e)p(e)} . For a rigorous implementation of this one would have to perform an exhaustive search by going through all strings e ∗ {\displaystyle e^{}} in the native language. Performing the search efficiently is the work of a machine translation decoder that uses the foreign string, heuristics and other methods to limit the search space and at the same time keeping acceptable quality. This trade-off between quality and time usage can also be found in speech recognition. As the translation systems are not able to store all native strings and their translations, a document is typically translated sentence by sentence. Language models are typically approximated by smoothed n-gram models, and similar approaches have been applied to translation models, but this introduces additional complexity due to different sentence lengths and word orders in the languages. Statistical translation models were initially word based (Models 1-5 from IBM Hidden Markov model from Stephan Vogel and Model 6 from Franz-Joseph Och), but significant advances were made with the introduction of phrase based models. Later work incorporated syntax or quasi-syntactic structures. == Benefits == The most frequently cited benefits of statistical machine translation (SMT) over rule-based approach are: More efficient use of human and data resources There are many parallel corpora in machine-readable format and even more monolingual data. Generally, SMT systems are not tailored to any specific pair of languages. More fluent translations owing to use of a language model == Shortcomings == Corpus creation can be costly. Specific errors are hard to predict and fix. Results may have superficial fluency that masks translation problems. Statistical machine translation usually works less well for language pairs with significantly different word order. The benefits obtained for translation between Western European languages are not representative of results for other language pairs, owing to smaller training corpora and greater grammatical differences. == Word-based translation == In word-based translation, the fundamental unit of translation is a word in some natural language. Typically, the number of words in translated sentences are different, because of compound words, morphology and idioms. The ratio of the lengths of sequences of translated words is called fertility, which tells how many foreign words each native word produces. Necessarily it is assumed by information theory that each covers the same concept. In practice this is not really true. For example, the English word corner can be translated in Spanish by either rincón or esquina, depending on whether it is to mean its internal or external angle. Simple word-based translation cannot translate between languages with different fertility. Word-based translation systems can relatively simply be made to cope with high fertility, such that they could map a single word to multiple words, but not the other way about. For example, if we were translating from English to French, each word in English could produce any number of French words— sometimes none at all. But there is no way to group two English words producing a single French word. An example of a word-based translation system is the freely available GIZA++ package (GPLed), which includes the training program for IBM models and HMM model and Model 6. The word-based translation is not widely used today; phrase-based systems are more common. Most phrase-based systems are still using GIZA++ to align the corpus. The alignments are used to extract phrases or deduce syntax rules. And matching words in bi-text is still a problem actively discussed in the community. Because of the predominance of GIZA++, there are now several distributed implementations of it online. == Phrase-based translation == In phrase-based translation, the aim is to reduce the restrictions of word-based translation by translating whole sequences of words, where the lengths may differ. The sequences of words are called blocks or phrases. These are typically not linguistic phrases, but phrasemes that were found using statistical methods from corpora. It has been shown that restricting the phrases to linguistic phrases (syntactically motivated groups of words, see syntactic categories) decreased the quality of translation. The chosen phrases are further mapped one-to-one based on a phrase translation table, and may be reordered. This table could be learnt based on word-alignment, or directly from a parallel corpus. The second model is trained using the expectation maximization algorithm, similarly to the word-based IBM model. == Syntax-based translation == Syntax-based translation is based on the idea of translating syntactic units, rather than single words or strings of words (as in phrase-based MT), i.e. (partial) parse trees of sentences/utterances. Until the 1990s, with advent of strong stochastic parsers, the statistical counterpart of the old idea of syntax-based translation did not take off. Examples of this approach include DOP-based MT and later synchronous context-free grammars. == Hierarchical phrase-based translation == Hierarchical phrase-based translation combines the phrase-based and syntax-based approaches to translation. It uses synchronous context-free grammar rules, but the grammars can be constructed by an extension of methods for phrase-based translation without reference to linguistically motivated syntactic constituents. This idea was first introduced in Chiang's Hiero system (2005). == Language models == A language model is an essential component of any statistical machine translation system, which aids in making the translation as fluent as possible. It is a function that takes a translated sentence and returns the probability of it being said by a native speaker. A good language model will for example assign a higher probability to the sentence "the house is small" than to "small the is house". Other than word order, language models may also help with word choice: if a foreign word has multiple possible translations, these functions may give better probabilities for certain translations in specific contexts in the target language. == Systems implementing statistical machine translation == Google Translate (started transition to neural machine translation in 2016) Microsoft Translator (started transition to neural machine translation in 2016) Yandex.Translate (switched to hybrid approach incorporating neural machine translation in 2017) == Challenges with statistical machine translation == Problems with statistical machine translation include: === Sentence alignment === Single sentences in one language can be found translated into several sentences in the o

  Read more →
• 

  András Kornai

  András Kornai (born 1957 in Budapest) is a mathematical linguist. == Education == Kornai is the son of economist János Kornai. He earned two PhDs with the first being in mathematics in 1983 from Eötvös Loránd University in Budapest, where his advisor was Miklós Ajtai. His second was in linguistics in 1991 from Stanford University, where his advisor was Paul Kiparsky. == Career == He is a professor in the Department of Algebra at the Budapest Institute of Technology, where he works on an open source Hungarian morphological analyzer. He was Chief Scientist at MetaCarta, where he worked on information extraction before the company was acquired by Nokia. Prior to MetaCarta, he was Chief Scientist at Northern Light. He is on the board of the journal Grammars and YourAmigo PLC. His research interests include all mathematical aspects of natural language processing, speech recognition, and OCR. As area editor he was responsible for the Mathematical Linguistics area of the Oxford International Encyclopedia of Linguistics, and his joint work with Geoffrey Pullum, "The X-bar Theory of Phrase Structure", formally reconstructed that then-popular linguistic theory. == Awards and honors == 2009: ACM Distinguished Member == Monographs == Semantics. Springer Nature, 2020. ISBN 978-3-319-65644-1 Mathematical Linguistics. Springer Verlag, in the series Advanced Information and Knowledge Processing, November 2007. ISBN 978-1-84628-985-9 Hardbound, approximately 300 pages. See description. Formal Phonology. In the series Outstanding Dissertations in Linguistics, Garland Publishing, 1994, ISBN 0-8153-1730-1, hardbound, 240 pages Contents, Preface, Introduction (20 pages) On Hungarian Morphology. In the series Linguistica, Hungarian Academy of Sciences, 1994, ISBN 963-8461-73-X, paperbound, 174 pages Contents, Preface, Introduction (10 pages) == Books edited == Oxford International Encyclopedia of Linguistics (Mathematical Linguistics Area Editor under Editor in Chief William Frawley). 4 volumes, Oxford University Press, 2003, ISBN 978-0-19-513977-8. Proceedings of the HLT-NAACL Workshop on the Analysis of Geographic References. Jointly with Beth Sundheim. Association for Computational Linguistics, 2003, ISBN 1-932432-04-3 (WS9), paperbound, vi+81 pages. See related material. Extended Finite State Models of Language (editor). In the series Studies in Natural Language Processing, Cambridge University Press, 1999, ISBN 0-521-63198-X, hardbound, x+278 pages Contents, Introduction (7 pages). == Selected papers == Digital Language Death. PLoS ONE 8(10): e77056, 2012. [1] Hunmorph: open source word analysis (Jointly with V. Tron, Gy. Gyepesi, P. Halacsy, L. Nemeth, and D. Varga). In Proc. ACL 2005 Software Workshop 77-85 [2] Leveraging the open source ispell codebase for minority language analysis (Jointly with P. Halacsy, L. Nemeth, A. Rung, I. Szakadat, and V. Tron). In J. Carson-Berndsen (ed): Proc. SALTMIL 2004 56-59 [3] Explicit Finitism, International Journal of Theoretical Physics 2003/2 301-307 [4] Mathematical Linguistics (Jointly with G.K. Pullum) In W. Frawley (ed): Oxford International Encyclopedia of Linguistics, Oxford University Press 2003, v3 17-20 [5] Optical Character Recognition, In W. Frawley (ed): Oxford International Encyclopedia of Linguistics, Oxford University Press 2003, v3 33-34 [6] How many words are there? Glottometrics 2002/4 61-86 [7] Zipf's law outside the middle range Proc. Sixth Meeting on Mathematics of Language University of Central Florida, 1999 347-356 [8] A Robust, Language-Independent OCR System. (Jointly with Z. Lu, I. Bazzi, J. Makhoul, P. Natarajan, and R. Schwartz) In: Robert J. Mericsko (ed): Proc. 27th AIPR Workshop: Advances in Computer-Assisted Recognition SPIE Proceedings 3584 1999 [9] Quantitative Comparison of Languages. Grammars 1998/2 155-165 [10] The generative power of feature geometry. Annals of Mathematics and Artificial Intelligence 8 1993 37-46 [11] The X-bar Theory of Phrase Structure. (Jointly with G.K. Pullum) Language 66 1990 24-50 [12]

  Read more →
• 

  Image analysis

  Image analysis or imagery analysis is the extraction of meaningful information from images; mainly from digital images by means of digital image processing techniques. Image analysis tasks can be as simple as reading bar coded tags or as sophisticated as identifying a person from their face. Computers are indispensable for the analysis of large amounts of data, for tasks that require complex computation, or for the extraction of quantitative information. On the other hand, the human visual cortex is an excellent image analysis apparatus, especially for extracting higher-level information, and for many applications — including medicine, security, and remote sensing — human analysts still cannot be replaced by computers. For this reason, many important image analysis tools such as edge detectors and neural networks are inspired by human visual perception models. == Digital == Digital Image Analysis or Computer Image Analysis is when a computer or electrical device automatically studies an image to obtain useful information from it. Note that the device is often a computer but may also be an electrical circuit, a digital camera or a mobile phone. It involves the fields of computer or machine vision, and medical imaging, and makes heavy use of pattern recognition, digital geometry, and signal processing. This field of computer science developed in the 1950s at academic institutions such as the MIT A.I. Lab, originally as a branch of artificial intelligence and robotics. It is the quantitative or qualitative characterization of two-dimensional (2D) or three-dimensional (3D) digital images. 2D images are, for example, to be analyzed in computer vision, and 3D images in medical imaging. The field was established in the 1950s—1970s, for example with pioneering contributions by Azriel Rosenfeld, Herbert Freeman, Jack E. Bresenham, or King-Sun Fu. == Techniques == There are many different techniques used in automatically analysing images. Each technique may be useful for a small range of tasks, however there still aren't any known methods of image analysis that are generic enough for wide ranges of tasks, compared to the abilities of a human's image analysing capabilities. Examples of image analysis techniques in different fields include: 2D and 3D object recognition, image segmentation, motion detection e.g. Single particle tracking, video tracking, optical flow, medical scan analysis, 3D Pose Estimation. == Deep learning == Since the early 2010s, deep learning methods have substantially advanced the field of image analysis. In 2012, a deep convolutional neural network (CNN) known as AlexNet achieved a significant reduction in error rates on the ImageNet large-scale image classification benchmark, demonstrating the effectiveness of deep learning for visual recognition tasks. Subsequent architectures such as ResNet introduced residual connections that enabled training of much deeper networks, further improving accuracy across image analysis tasks. Real-time object detection became practical with frameworks such as YOLO (You Only Look Once), which unified detection and classification into a single network pass. In 2020, the Vision Transformer (ViT) demonstrated that transformer architectures, originally developed for natural language processing, could achieve competitive results on image classification when applied directly to sequences of image patches. More recently, foundation models trained on large-scale datasets have enabled zero-shot generalisation across image analysis tasks. The Segment Anything Model (SAM), trained on over one billion masks, can segment arbitrary objects in images without task-specific fine-tuning. These advances have made image analysis techniques increasingly accessible through browser-based tools and open-source implementations. == Applications == The applications of digital image analysis are continuously expanding through all areas of science and industry, including: anatomy, allows for precise measurements, visualization, and statistical analysis of anatomical structures. assay micro plate reading, such as detecting where a chemical was manufactured. astronomy, such as calculating the size of a planet. automated species identification (e.g. plant and animal species) defense error level analysis filtering machine vision, such as to automatically count items in a factory conveyor belt. materials science, such as determining if a metal weld has cracks. medicine, such as detecting cancer in a mammography scan. metallography, such as determining the mineral content of a rock sample. microscopy, such as counting the germs in a swab. automatic number plate recognition; optical character recognition, such as automatic license plate detection. remote sensing, such as detecting intruders in a house, and producing land cover/land use maps. robotics, such as to avoid steering into an obstacle. security, such as detecting a person's eye color or hair color. == Object-based == Object-based image analysis (OBIA) involves two typical processes, segmentation and classification. Segmentation helps to group pixels into homogeneous objects. The objects typically correspond to individual features of interest, although over-segmentation or under-segmentation is very likely. Classification then can be performed at object levels, using various statistics of the objects as features in the classifier. Statistics can include geometry, context and texture of image objects. Over-segmentation is often preferred over under-segmentation when classifying high-resolution images. Object-based image analysis has been applied in many fields, such as cell biology, medicine, earth sciences, and remote sensing. For example, it can detect changes of cellular shapes in the process of cell differentiation.; it has also been widely used in the mapping community to generate land cover. When applied to earth images, OBIA is known as geographic object-based image analysis (GEOBIA), defined as "a sub-discipline of geoinformation science devoted to (...) partitioning remote sensing (RS) imagery into meaningful image-objects, and assessing their characteristics through spatial, spectral and temporal scale". The international GEOBIA conference has been held biannually since 2006. OBIA techniques are implemented in software such as eCognition or the Orfeo toolbox.

  Read more →
• 

  AI Writing Assistants Reviews: What Actually Works in 2026

  Looking for the best AI writing assistant? An AI writing assistant is software that uses machine learning to help you get more done — it can save you hours every week by automating repetitive work. Most options offer a generous free tier, with paid plans unlocking higher limits, faster processing, and team features. Whether you are a beginner or a pro, the right AI writing assistant slots into your workflow and pays for itself fast. Read on for hands-on impressions, pricing tiers, and the standout features that matter.

  Read more →
• 

  Hans Uszkoreit

  Hans Uszkoreit is a German computational linguist. Hans Uszkoreit studied Linguistics and Computer Science at Technische Universität Berlin and the University of Texas at Austin. While he was studying in Austin, he also worked as a research associate in a large machine translation project at the Linguistics Research Center. After he received his Ph.D. in linguistics from the University of Texas, he worked as a computer scientist at the Artificial Intelligence Center and was affiliated with the Center for the Study of Language and Information at Stanford University. Nowadays, he is teaching as a professor of Computational Linguistics at Saarland University. Moreover, he serves as a Scientific Director at the German Research Center for Artificial Intelligence (DFKI) where he heads the DFKI Language Technology Lab. == Life and career == Hans Uszkoreit, a native of East Berlin, was actively involved in a group of young individuals who opposed the East Germany regime. His protesting against the 1968 invasion of Czechoslovakia led to his expulsion from high school and subsequent imprisonment for a period of fifteen months on charges of subversive agitation. Realizing that continuing his education in East Germany was not feasible, Uszkoreit made the decision to escape to West Berlin. There, he completed his high school education and pursued a degree in Linguistics and Computer Science at Technische Universität Berlin. During his time as a student, he worked part-time as an editor and writer for Zitty, a city magazine, which he co-founded. In 1977, Uszkoreit was granted a Fulbright Grant to further his studies at the University of Texas at Austin. During his time in Austin, he concurrently served as a research associate in a significant machine translation project. Subsequently, he received a second Fulbright grant, which enabled him to pursue a Ph.D. program in linguistics. In 1984, he successfully completed his doctoral studies, earning a Ph.D. in linguistics. Between 1982 and 1986, Uszkoreit held the position of a computer scientist at the Artificial Intelligence Center of SRI International in Menlo Park, California. In 1988, he created the Department of Computational Linguistics and Phonetics at Saarland University. In 1989 he was elected head of the Language Technology Lab at DFKI. In 2012, Uszkoreit's achievements in the domain of relation extraction led to his receipt of a Google Faculty Research Award, acknowledging the substantial progress made by Uszkoreit and his team in advancing the field. In 2013, Uszkoreit, in collaboration with Feiyu Xu and Roberto Navigli, was granted an additional Google Research Award, which provided support for a targeted project within Google's Language Understanding Program, focusing on the augmentation of language comprehension and analysis. == Personal life == He is father of a son Jakob Uszkoreit, machine learning researcher scientist, an author of the landmark paper "Attention Is All You Need", and daughter Lena Uszkoreit. == Awards == 2002 Elected Member of the European Academy of Sciences 2012 Google Faculty Research Award 2013 Google Focused Research Award

  Read more →
• 

  Vector quantization

  Vector quantization (VQ) is a classical quantization technique from signal processing that allows the modeling of probability density functions by the distribution of prototype vectors. Developed in the early 1980s by Robert M. Gray, it was originally used for data compression. It works by dividing a large set of points (vectors) into groups having approximately the same number of points closest to them. Each group is represented by its centroid point, as in k-means and some other clustering algorithms. In simpler terms, vector quantization chooses a set of points to represent a larger set of points. The density matching property of vector quantization is powerful, especially for identifying the density of large and high-dimensional data. Since data points are represented by the index of their closest centroid, commonly occurring data have low error, and rare data high error. This is why VQ is suitable for lossy data compression. It can also be used for lossy data correction and density estimation. Vector quantization is based on the competitive learning paradigm, so it is closely related to the self-organizing map model and to sparse coding models used in deep learning algorithms such as autoencoder. == Training == One simple training algorithm for vector quantization is: Pick a sample point at random Move the nearest quantization vector centroid towards this sample point, by a small fraction of the distance Repeat A more sophisticated algorithm reduces the bias in the density matching estimation and ensures that all points are used, by including an extra sensitivity parameter: Increase each centroid's sensitivity s i {\displaystyle s_{i}} by a small amount Pick a sample point P {\displaystyle P} at random For each quantization vector centroid c i {\displaystyle c_{i}} , let d ( P , c i ) {\displaystyle d(P,c_{i})} denote the distance of P {\displaystyle P} and c i {\displaystyle c_{i}} Find the centroid c i {\displaystyle c_{i}} for which d ( P , c i ) − s i {\displaystyle d(P,c_{i})-s_{i}} is the smallest Move c i {\displaystyle c_{i}} towards P {\displaystyle P} by a small fraction of the distance Set s i {\displaystyle s_{i}} to zero Repeat It is desirable to use a cooling schedule to produce convergence: see Simulated annealing. Another simple method is LBG, which is based on k-means. The algorithm can be iteratively updated with "live" data, rather than by picking random points from a data set, but this will introduce some bias if the data are temporally correlated over many samples. == Applications == Vector quantization is used for lossy data compression, lossy data correction, pattern recognition, density estimation and clustering. Lossy data correction, or prediction, is used to recover data missing from some dimensions. It is done by finding the nearest group with the data dimensions available, then predicting the result based on the values for the missing dimensions, assuming that they will have the same value as the group's centroid. For density estimation, the area/volume that is closer to a particular centroid than to any other is inversely proportional to the density (due to the density matching property of the algorithm). === Use in data compression === Vector quantization, also called "block quantization" or "pattern matching quantization" is often used in lossy data compression. It works by encoding values from a multidimensional vector space into a finite set of values from a discrete subspace of lower dimension. A lower-space vector requires less storage space, so the data is compressed. Due to the density matching property of vector quantization, the compressed data has errors that are inversely proportional to density. The transformation is usually done by projection or by using a codebook. In some cases, a codebook can be also used to entropy code the discrete value in the same step, by generating a prefix coded variable-length encoded value as its output. The set of discrete amplitude levels is quantized jointly rather than each sample being quantized separately. Consider a k-dimensional vector [ x 1 , x 2 , . . . , x k ] {\displaystyle [x_{1},x_{2},...,x_{k}]} of amplitude levels. It is compressed by choosing the nearest matching vector from a set of n-dimensional vectors [ y 1 , y 2 , . . . , y n ] {\displaystyle [y_{1},y_{2},...,y_{n}]} , with n < k. All possible combinations of the n-dimensional vector [ y 1 , y 2 , . . . , y n ] {\displaystyle [y_{1},y_{2},...,y_{n}]} form the vector space to which all the quantized vectors belong. Only the index of the codeword in the codebook is sent instead of the quantized values. This conserves space and achieves more compression. Twin vector quantization (VQF) is part of the MPEG-4 standard dealing with time domain weighted interleaved vector quantization. === Video codecs based on vector quantization === Bink video Cinepak Daala is transform-based but uses pyramid vector quantization on transformed coefficients Digital Video Interactive: Production-Level Video and Real-Time Video Indeo Microsoft Video 1 QuickTime: Apple Video (RPZA) and Graphics Codec (SMC) Sorenson SVQ1 and SVQ3 Smacker video VQA format, used in many games The usage of video codecs based on vector quantization has declined significantly in favor of those based on motion compensated prediction combined with transform coding, e.g. those defined in MPEG standards, as the low decoding complexity of vector quantization has become less relevant. === Audio codecs based on vector quantization === AMR-WB+ CELP CELT (now part of Opus) is transform-based but uses pyramid vector quantization on transformed coefficients Codec 2 DTS G.729 iLBC Ogg Vorbis TwinVQ === Use in pattern recognition === VQ was also used in the eighties for speech and speaker recognition. Recently it has also been used for efficient nearest neighbor search and on-line signature recognition. In pattern recognition applications, one codebook is constructed for each class (each class being a user in biometric applications) using acoustic vectors of this user. In the testing phase the quantization distortion of a testing signal is worked out with the whole set of codebooks obtained in the training phase. The codebook that provides the smallest vector quantization distortion indicates the identified user. The main advantage of VQ in pattern recognition is its low computational burden when compared with other techniques such as dynamic time warping (DTW) and hidden Markov model (HMM). The main drawback when compared to DTW and HMM is that it does not take into account the temporal evolution of the signals (speech, signature, etc.) because all the vectors are mixed up. In order to overcome this problem a multi-section codebook approach has been proposed. The multi-section approach consists of modelling the signal with several sections (for instance, one codebook for the initial part, another one for the center and a last codebook for the ending part). === Use as clustering algorithm === As VQ is seeking for centroids as density points of nearby lying samples, it can be also directly used as a prototype-based clustering method: each centroid is then associated with one prototype. By aiming to minimize the expected squared quantization error and introducing a decreasing learning gain fulfilling the Robbins-Monro conditions, multiple iterations over the whole data set with a concrete but fixed number of prototypes converges to the solution of k-means clustering algorithm in an incremental manner. === Generative adversarial networks (GAN) === VQ has been used to quantize a feature representation layer in the discriminator of generative adversarial networks. The feature quantization (FQ) technique performs implicit feature matching. It improves the GAN training, and yields an improved performance on a variety of popular GAN models: BigGAN for image generation, StyleGAN for face synthesis, and U-GAT-IT for unsupervised image-to-image translation.

  Read more →
• 

  Charge-coupled device

  A charge-coupled device (CCD) is an integrated circuit containing an array of linked, or coupled, capacitors. Under the control of an external circuit, each capacitor can transfer its electric charge to a neighboring capacitor. CCD sensors are a major technology used in digital imaging. In a CCD image sensor, pixels are represented by p-doped metal–oxide–semiconductor (MOS) capacitors. These MOS capacitors, the basic building blocks of a CCD, are biased above the threshold for inversion when image acquisition begins, allowing the conversion of incoming photons into electron charges at the semiconductor-oxide interface; the CCD is then used to read out these charges. Although CCDs are not the only technology to allow for light detection, CCD image sensors are widely used in professional, medical, and scientific applications where high-quality image data are required. In applications with less exacting quality demands, such as consumer and professional digital cameras, active pixel sensors, also known as CMOS sensors (complementary MOS sensors), are generally used. However, the large quality advantage CCDs enjoyed early on has narrowed over time and since the late 2010s CMOS sensors are the dominant technology, having largely if not completely replaced CCD image sensors. == History == The basis for the CCD is the metal–oxide–semiconductor (MOS) structure, with MOS capacitors being the basic building blocks of a CCD, and a depleted MOS structure used as the photodetector in early CCD devices. In the late 1960s, Willard Boyle and George E. Smith at Bell Labs were researching MOS technology while working on semiconductor bubble memory. They realized that an electric charge was the analog of the magnetic bubble and that it could be stored on a tiny MOS capacitor. As it was fairly straightforward to fabricate a series of MOS capacitors in a row, they connected a suitable voltage to them so that the charge could be stepped along from one to the next. This led to the invention of the charge-coupled device by Boyle and Smith in 1969. They conceived of the design of what they termed, in their notebook, "Charge 'Bubble' Devices". The initial paper describing the concept in April 1970 listed possible uses as memory, a delay line, and an imaging device. The device could also be used as a shift register. The essence of the design was the ability to transfer charge along the surface of a semiconductor from one storage capacitor to the next. The first experimental device demonstrating the principle was a row of closely spaced metal squares on an oxidized silicon surface electrically accessed by wire bonds. It was demonstrated by Gil Amelio, Michael Francis Tompsett and George Smith in April 1970. This was the first experimental application of the CCD in image sensor technology, and used a depleted MOS structure as the photodetector. The first patent (U.S. patent 4,085,456) on the application of CCDs to imaging was assigned to Tompsett, who filed the application in 1971. The first working CCD made with integrated circuit technology was a simple 8-bit shift register, reported by Tompsett, Amelio and Smith in August 1970. This device had input and output circuits and was used to demonstrate its use as a shift register and as a crude eight pixel linear imaging device. Development of the device progressed at a rapid rate. By 1971, Bell researchers led by Michael Tompsett were able to capture images with simple linear devices. Several companies, including Fairchild Semiconductor, RCA and Texas Instruments, picked up on the invention and began development programs. Fairchild's effort, led by ex-Bell researcher Gil Amelio, was the first with commercial devices, and by 1974 had a linear 500-element device and a 2D 100 × 100 pixel device. Peter L. P. Dillon, a scientist at Kodak Research Labs, invented the first color CCD image sensor by overlaying a color filter array on this Fairchild 100 x 100 pixel Interline CCD starting in 1974. Steven Sasson, an electrical engineer working for the Kodak Apparatus Division, invented a digital still camera using this same Fairchild 100 × 100 CCD in 1975. The interline transfer (ILT) CCD device was proposed by L. Walsh and R. Dyck at Fairchild in 1973 to reduce smear and eliminate a mechanical shutter. To further reduce smear from bright light sources, the frame-interline-transfer (FIT) CCD architecture was developed by K. Horii, T. Kuroda and T. Kunii at Matsushita (now Panasonic) in 1981. The first KH-11 KENNEN reconnaissance satellite equipped with charge-coupled device array (800 × 800 pixels) technology for imaging was launched in December 1976. Under the leadership of Kazuo Iwama, Sony started a large development effort on CCDs involving a significant investment. Eventually, Sony managed to mass-produce CCDs for their camcorders. Before this happened, Iwama died in August 1982. Subsequently, a CCD chip was placed on his tombstone to acknowledge his contribution. The first mass-produced consumer CCD video camera, the CCD-G5, was released by Sony in 1983, based on a prototype developed by Yoshiaki Hagiwara in 1981. Early CCD sensors suffered from shutter lag. This was largely resolved with the invention of the pinned photodiode (PPD). It was invented by Nobukazu Teranishi, Hiromitsu Shiraki and Yasuo Ishihara at NEC in 1980. They recognized that lag can be eliminated if the signal carriers could be transferred from the photodiode to the CCD. This led to their invention of the pinned photodiode, a photodetector structure with low lag, low noise, high quantum efficiency and low dark current. It was first publicly reported by Teranishi and Ishihara with A. Kohono, E. Oda and K. Arai in 1982, with the addition of an anti-blooming structure. The new photodetector structure invented at NEC was given the name "pinned photodiode" (PPD) by B.C. Burkey at Kodak in 1984. In 1987, the PPD began to be incorporated into most CCD devices, becoming a fixture in consumer electronic video cameras and then digital still cameras. Since then, the PPD has been used in nearly all CCD sensors and then CMOS sensors. In January 2006, Boyle and Smith were awarded the National Academy of Engineering Charles Stark Draper Prize, and in 2009 they were awarded the Nobel Prize for Physics for their invention of the CCD concept. Michael Tompsett was awarded the 2010 National Medal of Technology and Innovation, for pioneering work and electronic technologies including the design and development of the first CCD imagers. He was also awarded the 2012 IEEE Edison Medal for "pioneering contributions to imaging devices including CCD Imagers, cameras and thermal imagers". == Basics of operation == In a CCD for capturing images, there is a photoactive region (an epitaxial layer of silicon), and a transmission region made out of a shift register (the CCD, properly speaking). An image is projected through a lens onto the capacitor array (the photoactive region), causing each capacitor to accumulate an electric charge proportional to the light intensity at that location. A one-dimensional array, used in line-scan cameras, captures a single slice of the image, whereas a two-dimensional array, used in video and still cameras, captures a two-dimensional picture corresponding to the scene projected onto the focal plane of the sensor. Once the array has been exposed to the image, a control circuit causes each capacitor to transfer its contents to its neighbor (operating as a shift register). The last capacitor in the array dumps its charge into a charge amplifier, which converts the charge into a voltage. By repeating this process, the controlling circuit converts the entire contents of the array in the semiconductor to a sequence of voltages. In a digital device, these voltages are then sampled, digitized, and usually stored in memory; in an analog device (such as an analog video camera), they are processed into a continuous analog signal (e.g. by feeding the output of the charge amplifier into a low-pass filter), which is then processed and fed out to other circuits for transmission, recording, or other processing. == Detailed physics of operation == === Charge generation === Before the MOS capacitors are exposed to light, they are biased into the depletion region; in n-channel CCDs, the silicon under the bias gate is slightly p-doped or intrinsic. The gate is then biased at a positive potential, above the threshold for strong inversion, which will eventually result in the creation of an n channel below the gate as in a MOSFET. However, it takes time to reach this thermal equilibrium: up to hours in high-end scientific cameras cooled at low temperature. Initially after biasing, the holes are pushed far into the substrate, and no mobile electrons are at or near the surface; the CCD thus operates in a non-equilibrium state called deep depletion. Then, when electron–hole pairs are generated in the depletion region, they are separated by the electric field, the elec

  Read more →
• 

  Moses (machine translation)

  Moses is a statistical machine translation engine that can be used to train statistical models of text translation from a source language to a target language, developed by the University of Edinburgh. Moses then allows new source-language text to be decoded using these models to produce automatic translations in the target language. Training requires a parallel corpus of passages in the two languages, typically manually translated sentence pairs. Moses is free and open-source software, released under the GNU Library Public License (LGPL), and available as source code and binary files for Windows and Linux. Its development is supported mainly by the EuroMatrix project, with funding by the European Commission. Among its features are: A beam search algorithm that quickly finds the highest probability translation within a set of choices Phrase-based translation of short text chunks Handles words with multiple factored representations to enable integrating linguistic and other information (e.g., surface form, lemma and morphology, part-of-speech, word class) Decodes ambiguous forms of a source sentence, represented as a confusion network, to support integrating with upstream tools such as speech recognizers Support for large language models (LMs) such as IRSTLM (an exact LM using memory-mapping) and RandLM (an inexact LM based on Bloom filters)

  Read more →
• 

  SYSTRAN

  SYSTRAN, founded by Dr. Peter Toma in 1968, is one of the oldest machine translation companies. SYSTRAN has done extensive work for the United States Department of Defense and the European Commission. SYSTRAN provided the technology for Yahoo! Babel Fish until May 30, 2012, among others. It was used by Google's language tools until 2007. SYSTRAN is used by the Dashboard Translation widget in macOS. Commercial versions of SYSTRAN can run on Microsoft Windows (including Windows Mobile), Linux, and Solaris. Historically, SYSTRAN systems used rule-based machine translation (RbMT) technology. With the release of SYSTRAN Server 7 in 2010, SYSTRAN implemented a hybrid rule-based/statistical machine translation (SMT) technology which was the first of its kind in the marketplace. As of 2008, the company had 59 employees of whom 26 are computational experts and 15 computational linguists. The number of employees decreased from 70 in 2006 to 59 in 2008. In January 2024, ChapsVision acquired Systran. == History == With its origin in the Georgetown machine translation effort, SYSTRAN was one of the few machine translation systems to survive the major decrease of funding after the ALPAC Report of the mid-1960s. The company was established in La Jolla in California to work on translation of Russian to English text for the United States Air Force during the Cold War. Large numbers of Russian scientific and technical documents were translated using SYSTRAN under the auspices of the USAF Foreign Technology Division (later the National Air and Space Intelligence Center) at Wright-Patterson Air Force Base, Ohio. The quality of the translations, although only approximate, was usually adequate for understanding content. The company headquarters is in Paris, while its U.S. headquarters is in San Diego, CA. During the dot-com boom, the international language industry started a new era, and SYSTRAN entered into agreements with a number of translation integrators, the most successful of these being WorldLingo. In 2016, the Harvard NLP group and SYSTRAN founded OpenNMT, an open source ecosystem for neural machine translation and neural sequence learning. This has enabled machine translation software with learning capabilities, dramatically increasing MT translation quality. The project has since been used in several research and industry applications, and its open source ecosystem is currently maintained by SYSTRAN and Ubiqus. == Business situation == Most of SYSTRAN's revenue comes from a few customers. 57.1% comes from the 10 main customers and the three largest customers account for 10.9%, 8.9%, and 8.9% of its revenues, respectively. Revenues had been declining in the early 2000s: 10.2 million euros in 2004, 10.1 million euros in 2005, 9.3 million euros in 2006, 8.8 million euros in 2007, and 7.6 million euros in 2008, before seeing a rebound in 2009 with 8.6 million euros. == Languages == The following is a list of the languages in which SYSTRAN translate from and to English: Russian into English in 1968 and English into Russian in 1973 for the Apollo–Soyuz project.

  Read more →
• 

  Peter Gerstoft

  Peter Gerstoft is a Danish-American scientist and engineer specializing in ocean acoustics, seismology, and signal processing. He is currently a professor in the Department of Electrical and Photonics Engineering at the Technical University of Denmark. He was previously a Distinguished Data Scientist at the Scripps Institution of Oceanography at the University of California, San Diego and an adjunct professor in the Department of Electrical and Computer Engineering at UC San Diego. == Education == Gerstoft received his MSc in engineering from the Technical University of Denmark in 1983 and another MSc from the University of Western Ontario in 1984. He completed his PhD in engineering at the Technical University of Denmark in 1986. == Career == Gerstoft began his career in acoustics and vibrations at Odegaard & Danneskiold-Samsøe (1987–1992). He then served as a Senior Scientist at the NATO SACLANT Undersea Research Centre in La Spezia, Italy, from 1992 to 1997. Between 1999 and 2000, Gerstoft worked as a Senior Seismic Acoustic Officer with the Comprehensive Nuclear-Test-Ban Treaty Organization. He has been a Data Scientist at the Scripps Institution of Oceanography since 1997. From 2013, he held an adjunct faculty position in Electrical and Computer Engineering at UC San Diego, where he taught courses on seismology, data assimilation, and machine learning for physical systems. Gerstoft retired from UC San Diego in 2025 and accepted an appointment as Professor of Electrical and Photonics Engineering at the Technical University of Denmark in 2026 . == Research and contributions == Gerstoft's research focuses on environmental signal processing, with a particular emphasis on inversion methods, including their theoretical development, algorithmic implementation, and practical applications. In the 1990s, he investigated the use of nonlinear optimization and Bayesian approaches in acoustic inverse problems related to source localization and environmental parameter estimation. His work integrated physical propagation models with Bayesian sampling methods and a range of likelihood functions. These techniques have been applied to various data types, including vertical sensor arrays, single-sensor broadband data, and transmission loss measurements, and contributed to a general framework for inversion based on Gaussian assumptions. He has also conducted research in machine learning and sparse signal processing, particularly in the context of sensor array data. This includes applications such as direction of arrival estimation and source localization, including for seismic events such as the 2011 Tōhoku earthquake and for ship tracking in ocean environments. His work on sparse Bayesian sequential methods and techniques for estimating Lagrange multipliers in constrained optimization problems has contributed to the development of adaptive and high-resolution signal processing techniques. Gerstoft has applied supervised learning and deep neural networks to problems in physical acoustics, including source localization in ocean waveguides. He has also co-authored several review articles on the use of machine learning in acoustics and seismology. == Honors == Fulbright Scholar, Massachusetts Institute of Technology (1989–1990) Fellow, Acoustical Society of America (2003) Member, American Geophysical Union (since 2004) Senior Member, Institute of Electrical and Electronics Engineers (2018) Fellow, Institute of Electrical and Electronics Engineers (2023) == Selected publications == === Book === Diachok, O., Caiti, A., Gerstoft, P., & Schmidt, H. (Eds.). Full Field Inversion Methods in Ocean and Seismo-Acoustics. Kluwer Academic Publishers, 1995. === Selected articles === Gerstoft, P. (1994). "Inversion of seismo-acoustic data using genetic algorithms and a posteriori probability distributions". Journal of the Acoustical Society of America. 95 (2): 770–782. doi:10.1121/1.408467. Gerstoft, P., & Mecklenbrauker, C. F. (1998). "Ocean acoustic inversion with estimation of a posteriori probability distributions". Journal of the Acoustical Society of America. 104 (2): 808–819. doi:10.1121/1.423287. Sabra, K. G., Gerstoft, P., Roux, P., Kuperman, W. A., & Fehler, M. (2005). "Extracting time-domain Green's function estimates from ambient seismic noise". Geophysical Research Letters. 32, L03310. Xenaki, A., Gerstoft, P., & Mosegaard, K. (2014). "Compressive beamforming". Journal of the Acoustical Society of America. 136, 260–271. Niu, H., Reeves, D., & Gerstoft, P. (2017). "Source localization in an ocean waveguide using supervised machine learning". Journal of the Acoustical Society of America. 142, 1176–1188.

  Read more →
• 

  ViEWER

  ViEWER, the Virtual Environment Workbench for Education and Research, is a proprietary, freeware computer program for Microsoft Windows written by researchers at the University of Idaho for the study of visual perception and complex immersive three-dimensional environments. It was created using C++ and OpenGL, and has been used by Dr. Brian Dyre, Dr. Steffen Werner, Dr. Ernesto Bustamante, Dr. Ben Barton, and their undergraduate and graduate researchers in visual perception, signal detection, and child-safety experiments.

  Read more →
• 

  Bob Coecke

  Bob Coecke (born 23 July 1968) is a Belgian theoretical physicist and logician. He was Professor of Quantum foundations, Logics, and Structures at Oxford University until 2020. He was Chief Scientist at quantum computing company Quantinuum, until 2025 and founded a startup called Relational Intelligence in 2026. He is also Distinguished Visiting Research Chair at the Perimeter Institute for Theoretical Physics, and Emeritus Fellow at Wolfson College, Oxford. He pioneered categorical quantum mechanics (entry 18M40 in Mathematics Subject Classification 2020), Quantum Picturalism, ZX-calculus, DisCoCat model for natural language,, quantum natural language processing (QNLP) and quantum education through the book Quantum in Pictures. He is a founder of the Quantum Physics and Logic community and the Applied Category Theory communities and conference series, and of the journal Compositionality. Coecke is also a composer and musician, who has been called a pioneer of industrial music, and is also one of the pioneers of employing quantum computers in music. == Education and career == Coecke obtained his doctorate in sciences at the Vrije Universiteit Brussel in 1996, and performed postdoctoral work in the Theoretical Physics Group of Imperial College, London in the Category Theory Group of the Mathematics and Statistics Department at McGill University in Montreal, in the Department of Pure Mathematics and Mathematical Statistics of Cambridge University, and in the Department of Computer Science, University of Oxford. He was an EPSRC Advanced Research Fellow at the Department of Computer Science, University of Oxford, where he became Lecturer in Quantum Computer Science in 2007, and jointly with Samson Abramsky built and headed the Quantum Group. In July 2011, he was nominated professor of Quantum Foundations, Logics and Structures at Oxford University, with retroactive effect as of October 2010. He was a Governing Body Fellow of Wolfson College, Oxford since 2007, where he now is an Emeritus Fellow. In January 2019, Coecke became Senior Scientific Advisor of Cambridge Quantum Computing, and in January 2021 he resigned from his Professorship at Oxford, to become Chief Scientist of Cambridge Quantum Computing. After the merger of Cambridge Quantum Computing with Honeywell Quantum Systems, he stayed on as Chief Scientist of the joint entity Quantinuum until 2025. In January 2023 he also became Distinguished Visiting Research Chair at the Perimeter Institute for Theoretical Physics. == Work == Coecke's research focuses on the foundations of physics, more particularly category theory, logic, and diagrammatic reasoning, with application to quantum informatics, quantum gravity, and NLP. He has pioneered categorical quantum mechanics together with Samson Abramsky, and spearheaded the development of a diagrammatic quantum formalism based on Penrose graphical notation, on which he wrote a textbook entitled Picturing Quantum Processes with Aleks Kissinger. With Ross Duncan he pioneered ZX-calculus. He pioneered the DisCoCat model for natural language, with Stephen Clark and Mehrnoosh Sadrzadeh. He also pioneered quantum natural language processing (QNLP), with Will Zeng, and colleagues at Cambridge Quantum Computing. == Music == Coecke is also a musician, performing and recording since the eighties. He retrospectively has been named a pioneer of industrial music. His band, Black Tish, "used cutting edge sampling techniques for the time, a host of synth and sound loops and metal-style guitars to create a heavy rock/electronica fusion unlike anything heard before", and "bridge the gap between the pure experimental nature of bands like Throbbing Gristle and Einstürzende Neubauten and the (comparatively) more radio accessible Ministry or Nine Inch Nails". Coecke is also one of the pioneers of employing quantum computers in music. == Selected publications == Textbooks Bob Coecke, Aleks Kissinger:Picturing Quantum Processes. A First Course in Quantum Theory and Diagrammatic Reasoning, Cambridge University Press, 2017, ISBN 978-1316219317 Bob Coecke, Stefano Gogioso:Quantum in Pictures, Quantinuum, 2022, ISBN 978-1-7392147-1-5 Books (as editor) Bob Coecke, David Moore, Alexander Wilce (eds.): Current Research in Operational Quantum Logic: Algebras, Categories, Languages, Fundamental Theories of Physics, Kluwer Academic, 2010, ISBN 978-9048154371 Bob Coecke (ed.): New Structures for Physics, Lecture Notes in Physics 813, Springer, 2011, ISBN 978-3642128202 Articles Bob Coecke: Kindergarten quantum mechanics, arXiv:quant-ph/0510032 Samson Abramsky, Bob Coecke: A categorical semantics of quantum protocols, Proceedings of the 19th Annual IEEE Symposium on Logic in Computer Science, 2004, pp. 415–425 Bob Coecke, Ross Duncan: Interacting quantum observables, Automata, Languages and Programming, pp. 298–310, 2008 Konstantinos Meichanetzidis, Alexis Toumi, Giovanni de Felice, Bob Coecke: Grammar-Aware Question-Answering on Quantum Computers, arXiv:2012.03756 Bob Coecke: The Mathematics of Text Structure, arXiv:1904.03478 Will Zeng, Bob Coecke: Quantum Algorithms for Compositional Natural Language Processing, arXiv:1608.01406 Bob Coecke, Tobias Fritz, Robert Spekkens: A mathematical theory of resources, arXiv:1409.5531 Bob Coecke: An Alternative Gospel of structure: order, composition, processes, arxiv:1307.4038 Bob Coecke, Mehrnoosh Sadrzadeh, Steven Clark: Mathematical Foundations for a Compositional Distributional Model of Meaning, arXiv:1003.4394 Bob Coecke: Quantum Picturalism, arXiv:0908.1787 Software articles Eduardo Reck Miranda, Richie Yeung, Anna Pearson, Konstantinos Meichanetzidis, Bob Coecke: A quantum natural language processing approach to musical intelligence, arXiv:2111.06741 Dimitri Kartsaklis, Ian Fan, Richie Yeung, Anna Pearson, Robin Lorenz, Alexis Toumi, Giovanni de Felice, Konstantinos Meichanetzidis, Stephen Clark, Bob Coecke: lambeq: An efficient high-level python library for quantum NLP, arXiv:2110.04236 Giovanni de Felice, Alexis Toumi, Bob Coecke: Discopy: monoidal categories in Python, arXiv:2111.06741

  Read more →
• 

  Hidden Markov model

  A hidden Markov model (HMM) is a Markov model in which the observations are dependent on a latent (or hidden) Markov process (referred to as X {\displaystyle X} ). An HMM requires that there be an observable process Y {\displaystyle Y} whose outcomes depend on the outcomes of X {\displaystyle X} in a known way. Since X {\displaystyle X} cannot be observed directly, the goal is to learn about state of X {\displaystyle X} by observing Y {\displaystyle Y} . By definition of being a Markov model, an HMM has an additional requirement that the outcome of Y {\displaystyle Y} at time t = t 0 {\displaystyle t=t_{0}} must be "influenced" exclusively by the outcome of X {\displaystyle X} at t = t 0 {\displaystyle t=t_{0}} and that the outcomes of X {\displaystyle X} and Y {\displaystyle Y} at t < t 0 {\displaystyle t