AI Assistant Zia Spark Icon

AI Assistant Zia Spark Icon — independent reviews, comparisons, pricing and step-by-step guides on Aizhi.

• 

  Scanimate

  Scanimate is an analog computer animation (video synthesizer) system created by Lee Harrison III of Denver, Colorado. Harrison had developed its predecessor, ANIMAC, which generated used a motion capture system, based on a body suit with potentiometers. In contrast, Scanimate included TV technology. Scanimate's successor was called Caesar, and used a digital computer to control the analog system. The eight Scanimate systems were used to produce much of the video-based animation seen on television between most of the 1970s and early 1980s in commercials, promotions, and show openings. One of the major advantages the Scanimate system had over film-based animation and computer animation was the ability to create animations in real time. The speed with which animation could be produced on the system because of this, as well as its range of possible effects, helped it to supersede film-based animation techniques for television graphics. By the mid-1980s, it was superseded by digital computer animation, which produced sharper images and more sophisticated 3D imagery. Animations created on Scanimate and similar analog computer animation systems have a number of characteristic features that distinguish them from film-based animation: the motion is extremely fluid, using all 60 fields per second (in NTSC format video) or 50 fields (in PAL format video) rather than the 24 frames per second that film uses; the colors are much brighter and more saturated; and the images have a very "electronic" look that results from the direct manipulation of video signals through which the Scanimate produces the images. == How it works == A special high-resolution (around 945 lines) monochrome camera records high-contrast artwork. The image is then displayed on a high-resolution screen. Unlike a normal monitor, its deflection signals are passed through a special analog computer that enables the operator to bend the image in a variety of ways. The image is then shot from the screen by either a film camera or a video camera. In the case of a video camera, this signal is then fed into a colorizer, a device that takes certain shades of grey and turns it into color as well as transparency. The idea behind this is that the output of the Scanimate itself is always monochrome. Another advantage of the colorizer is that it gives the operator the ability to continuously add layers of graphics. This makes possible the creation of very complex graphics. This is done by using two video recorders. The background is played by one recorder and then recorded by another one. This process is repeated for every layer. This requires very high-quality video recorders (such as both the Ampex VR-2000 or IVC's IVC-9000 of Scanimate's era, the IVC-9000 being used quite frequently for Scanimate composition due to its very high generational quality between re-recordings). == Current usage == Two of the Scanimates are still in use at ZFx studios in Asheville, North Carolina. The original "Black Swan" R&D machine has been updated with more modern power supplies and can produce material in standard or 1080P high definition video. The "white Pearl" machine is the last one produced and is being kept in its original configuration for historical purposes by David Sieg at ZFx inc. The machines are installed in a working production environment with Grass Valley switchers, Kaleidoscope digital video effects systems, and Accom digital disk recorders for layering. == Use in television, music and films == === Music videos === Let's Groove by Earth, Wind & Fire Get Down on It by Kool & the Gang Blame It on the Boogie by The Jacksons Knock on Wood by Amii Stewart Popcorn Love by New Edition === TV programs/movies === === TV channels/home video/TV productions ===

  Read more →
• 

  Best AI Essay Writers in 2026

  Comparing the best AI essay writer? An AI essay writer is software that uses machine learning to help you get more done — it lowers the barrier so anyone can produce professional output. Privacy matters too: check whether your data trains the model and whether a no-log or enterprise tier is available. Whether you are a beginner or a pro, the right AI essay writer slots into your workflow and pays for itself fast. We tested the leading options and ranked them by quality, value, and ease of use.

  Read more →
• 

  AI Content Generators: Free vs Paid (2026)

  Shopping for the best AI content generator? An AI content generator is software that uses machine learning to help you get more done — it keeps getting smarter as the underlying models improve. Pricing, accuracy, and the size of the model behind the tool are the three factors that most affect daily usefulness. Whether you are a beginner or a pro, the right AI content generator slots into your workflow and pays for itself fast. We tested the leading options and ranked them by quality, value, and ease of use.

  Read more →
• 

  Laws of Form

  Laws of Form (hereinafter LoF) is a book by G. Spencer-Brown, written by August 1967 and published in 1969. The book straddles the boundary between mathematics and philosophy. LoF describes three distinct logical systems: The primary arithmetic (described in Chapter 4 of LoF), whose models include Boolean arithmetic; The primary algebra (Chapter 6 of LoF), whose models include the two-element Boolean algebra (hereinafter abbreviated 2), Boolean logic, and the classical propositional calculus; Equations of the second degree (Chapter 11), whose interpretations include finite automata and Alonzo Church's Restricted Recursive Arithmetic (RRA). "Boundary algebra" is a Meguire (2011) term for the union of the primary algebra and the primary arithmetic. Laws of Form sometimes loosely refers to the "primary algebra" as well as to LoF. == Contents == The preface states that the work was first explored in 1959, and Spencer Brown cites Bertrand Russell as being supportive of his endeavour. He also thanks J. C. P. Miller of University College London for helping with the proofreading and offering other guidance. In 1963 Spencer Brown was invited by Harry Frost, staff lecturer in the physical sciences at the department of Extra-Mural Studies of the University of London, to deliver a course on the mathematics of logic. LoF emerged from work in electronic engineering its author did around 1960. Key ideas of the LOF were first outlined in his 1961 manuscript Design with the Nor, which remained unpublished until 2021, and further refined during subsequent lectures on mathematical logic he gave under the auspices of the University of London's extension program. LoF has appeared in several editions. The second series of editions appeared in 1972 with the "Preface to the First American Edition", which emphasised the use of self-referential paradoxes, and the most recent being a 1997 German translation. LoF has never gone out of print. LoF's mystical and declamatory prose and its love of paradox make it a challenging read for all. Spencer-Brown was influenced by Ludwig Wittgenstein and R. D. Laing. LoF also echoes a number of themes from the writings of Charles Sanders Peirce, Bertrand Russell, and Alfred North Whitehead. The work has had curious effects on some classes of its readership; for example, on obscure grounds, it has been claimed that the entire book is written in an operational way, giving instructions to the reader instead of telling them what "is", and that in accordance with G. Spencer-Brown's interest in paradoxes, the only sentence that makes a statement that something is, is the statement which says no such statements are used in this book. Furthermore, the claim asserts that except for this one sentence the book can be seen as an example of E-Prime. What prompted such a claim, is obscure, either in terms of incentive, logical merit, or as a matter of fact, because the book routinely and naturally uses the verb to be throughout, and in all its grammatical forms, as may be seen both in the original and in quotes shown below. == Reception == Ostensibly a work of formal mathematics and philosophy, LoF became something of a cult classic: it was praised by Heinz von Foerster when he reviewed it for the Whole Earth Catalog. Those who agree point to LoF as embodying an enigmatic "mathematics of consciousness", its algebraic symbolism capturing an (perhaps even "the") implicit root of cognition: the ability to "distinguish". LoF argues that primary algebra reveals striking connections among logic, Boolean algebra, and arithmetic, and the philosophy of language and mind. Stafford Beer wrote in a review for Nature in 1969, "When one thinks of all that Russell went through sixty years ago, to write the Principia, and all we his readers underwent in wrestling with those three vast volumes, it is almost sad". Banaschewski (1977) argues that the primary algebra is nothing but new notation for Boolean algebra. Indeed, the two-element Boolean algebra 2 can be seen as the intended interpretation of the primary algebra. Yet the notation of the primary algebra: Fully exploits the duality characterizing not just Boolean algebras but all lattices; Highlights how syntactically distinct statements in logic and 2 can have identical semantics; Dramatically simplifies Boolean algebra calculations, and proofs in sentential and syllogistic logic. Moreover, the syntax of the primary algebra can be extended to formal systems other than 2 and sentential logic, resulting in boundary mathematics. LoF has influenced, among others, Heinz von Foerster, Louis Kauffman, Niklas Luhmann, Humberto Maturana, Francisco Varela and William Bricken. Some of these authors have modified the primary algebra in a variety of interesting ways. LoF claimed that certain well-known mathematical conjectures of very long standing, such as the four color theorem, Fermat's Last Theorem, and the Goldbach conjecture, are provable using extensions of the primary algebra. Spencer-Brown eventually circulated a purported proof of the four color theorem, but it was met with skepticism. == The form (Chapter 1) == The symbol: Also called the "mark" or "cross", is the essential feature of the Laws of Form. In Spencer-Brown's inimitable and enigmatic fashion, the Mark symbolizes the root of cognition, i.e., the dualistic Mark indicates the capability of differentiating a "this" from "everything else but this". In LoF, a Cross denotes the drawing of a "distinction", and can be thought of as signifying the following, all at once: The act of drawing a boundary around something, thus separating it from everything else; That which becomes distinct from everything by drawing the boundary; Crossing from one side of the boundary to the other. All three ways imply an action on the part of the cognitive entity (e.g., person) making the distinction. As LoF puts it: "The first command: Draw a distinction can well be expressed in such ways as: Let there be a distinction, Find a distinction, See a distinction, Describe a distinction, Define a distinction, Or: Let a distinction be drawn". (LoF, Notes to chapter 2) The counterpoint to the Marked state is the Unmarked state, which is simply nothing, the void, or the un-expressable infinite represented by a blank space. It is simply the absence of a Cross. No distinction has been made and nothing has been crossed. The Marked state and the void are the two primitive values of the Laws of Form. The Cross can be seen as denoting the distinction between two states, one "considered as a symbol" and another not so considered. From this fact arises a curious resonance with some theories of consciousness and language. Paradoxically, the Form is at once Observer and Observed, and is also the creative act of making an observation. LoF (excluding back matter) closes with the words: ...the first distinction, the Mark and the observer are not only interchangeable, but, in the form, identical. C. S. Peirce came to a related insight in the 1890s; see § Related work. == The primary arithmetic (Chapter 4) == The syntax of the primary arithmetic goes as follows. There are just two atomic expressions: The empty Cross ; All or part of the blank page (the "void"). There are two inductive rules: A Cross may be written over any expression; Any two expressions may be concatenated. The semantics of the primary arithmetic are perhaps nothing more than the sole explicit definition in LoF: "Distinction is perfect continence". Let the "unmarked state" be a synonym for the void. Let an empty Cross denote the "marked state". To cross is to move from one value, the unmarked or marked state, to the other. We can now state the "arithmetical" axioms A1 and A2, which ground the primary arithmetic (and hence all of the Laws of Form): "A1. The law of Calling". Calling twice from a state is indistinguishable from calling once. To make a distinction twice has the same effect as making it once. For example, saying "Let there be light" and then saying "Let there be light" again, is the same as saying it once. Formally: = {\displaystyle \ =} "A2. The law of Crossing". After crossing from the unmarked to the marked state, crossing again ("recrossing") starting from the marked state returns one to the unmarked state. Hence recrossing annuls crossing. Formally: = {\displaystyle \ =} In both A1 and A2, the expression to the right of '=' has fewer symbols than the expression to the left of '='. This suggests that every primary arithmetic expression can, by repeated application of A1 and A2, be simplified to one of two states: the marked or the unmarked state. This is indeed the case, and the result is the expression's "simplification". The two fundamental metatheorems of the primary arithmetic state that: Every finite expression has a unique simplification. (T3 in LoF); Starting from an initial marked or unmarked state, "complicating" an expression by a finite number of repeated application of A1 and A2 cannot yield

  Read more →
• 

  Machine learning

  Machine learning (ML) is a field of study in artificial intelligence concerned with the development and study of statistical algorithms that can learn from data and generalize to unseen data, and thus perform tasks without being explicitly programmed. Advances in the field of deep learning have allowed neural networks, a class of statistical algorithms, to surpass many previous machine learning approaches in performance. Statistics and mathematical optimisation methods compose the foundations of machine learning. Data mining is a related field of study, focusing on exploratory data analysis (EDA) through unsupervised learning. From a theoretical viewpoint, probably approximately correct learning provides a mathematical and statistical framework for describing machine learning. Most traditional machine learning and deep learning algorithms can be described as empirical risk minimisation under this framework. == History == The term machine learning was coined in 1959 by Arthur Samuel, an IBM employee and pioneer in the field of computer gaming and artificial intelligence. The synonym self-teaching computers was also used during this time period. The earliest machine learning program was introduced in the 1950s, when Samuel invented a computer program that calculated the chance of winning in checkers for each side, but the history of machine learning is rooted in decades of efforts to study human cognitive processes. In 1949, Canadian psychologist Donald Hebb published the book The Organization of Behavior, in which he introduced a theoretical neural structure formed by certain interactions among nerve cells. The Hebbian theory of neuron interaction set the groundwork for how many machine learning algorithms work, with connected artificial neurons changing the strength of their connections based on data. Other researchers who have studied human cognitive systems contributed to the modern machine learning technologies as well, including Walter Pitts and Warren McCulloch, who proposed the first mathematical model of neural networks including algorithms that mirror human thought processes. By the early 1960s, an experimental "learning machine" with punched tape memory, called Cybertron, had been developed by Raytheon Company to analyse sonar signals, electrocardiograms, and speech patterns using rudimentary reinforcement learning. It was repetitively "trained" by a human operator/teacher to recognise patterns and equipped with a "goof" button to cause it to reevaluate incorrect decisions. A representative book on research into machine learning during the 1960s was Nils Nilsson's book "Learning Machines", dealing mostly with machine learning for pattern classification. Interest related to pattern recognition continued into the 1970s, as described by Duda and Hart in 1973. In 1981, a report was given on using teaching strategies so that an artificial neural network learns to recognise 40 characters (26 letters, 10 digits, and 4 special symbols) from a computer terminal. Tom M. Mitchell provided a widely quoted, more formal definition of the algorithms studied in the machine learning field: "A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P if its performance at tasks in T, as measured by P, improves with experience E." This definition of the tasks in which machine learning is concerned is fundamentally operational rather than defining the field in cognitive terms. This follows Alan Turing's proposal in his paper "Computing Machinery and Intelligence", in which the question, "Can machines think?", is replaced by asking whether machines can convincingly imitate a human in its responses to human-posed questions. In 2014 Ian Goodfellow and others introduced generative adversarial networks (GANs) which could produce realistic synthetic data. By 2016 AlphaGo had won against top human players in Go using reinforcement learning techniques. == Relationships to other fields == === Artificial intelligence === As a scientific endeavour, machine learning grew out of the quest for artificial intelligence (AI). In the early days of AI as an academic discipline, some researchers were interested in having machines learn from data. They attempted to approach the problem with various symbolic methods, as well as what were then termed "neural networks"; these were mostly perceptrons and other models that were later found to be reinventions of the generalised linear models of statistics. Probabilistic reasoning was also employed, especially in automated medical diagnosis. However, an increasing emphasis on the logical, knowledge-based approach caused a rift between AI and machine learning. Probabilistic systems were plagued by theoretical and practical problems of data acquisition and representation. By 1980, expert systems had come to dominate AI, and statistics was out of favour. Work on symbolic/knowledge-based learning continued within AI, leading to inductive logic programming (ILP), but the more statistical line of research was now outside the field of AI proper, in pattern recognition and information retrieval. Neural network research was abandoned by AI and computer science around the same time. This subfield, termed "connectionism", was continued by researchers from other disciplines, including John Hopfield, David Rumelhart, and Geoffrey Hinton. Their main success came in the mid-1980s with the reinvention of backpropagation. Machine learning (ML), reorganised and recognised as its own field, started to flourish in the 1990s. The field changed its goal from achieving artificial intelligence to tackling solvable problems of a practical nature. It shifted focus away from the symbolic approaches it had inherited from AI, and toward methods and models borrowed from statistics, fuzzy logic, and probability theory. === Data compression === === Data mining === Machine learning and data mining often employ the same methods and overlap significantly, but while machine learning focuses on prediction based on known properties learned from the training data, data mining focuses on the discovery of previously unknown properties in the data (this is the analysis step of knowledge discovery in databases). Data mining uses many machine learning methods, but with different goals; on the other hand, machine learning also employs data mining methods as "unsupervised learning" or as a preprocessing step to improve learner accuracy. Much of the confusion between these two research communities comes from the basic assumptions they work with: in machine learning, performance is usually evaluated with respect to the ability to reproduce known knowledge, while in knowledge discovery and data mining (KDD) the key task is the discovery of previously unknown knowledge. Evaluated with respect to known knowledge, an uninformed (unsupervised) method will easily be outperformed by other supervised methods, while in a typical KDD task, supervised methods cannot be used due to the unavailability of training data. Machine learning also has intimate ties to optimization: Many learning problems are formulated as minimisation of some loss function on a training set of examples. Loss functions express the discrepancy between the predictions of the model being trained and the actual problem instances (for example, in classification, one wants to assign a label to instances, and models are trained to correctly predict the preassigned labels of a set of examples). === Generalization === Characterizing the generalisation of various learning algorithms is an active topic of current research, especially for deep learning algorithms. === Statistics === Machine learning and statistics are closely related fields in terms of methods, but distinct in their principal goal: statistics draws population inferences from a sample, while machine learning finds generalisable predictive patterns. Conventional statistical analyses require the a priori selection of a model most suitable for the study data set. In addition, only significant or theoretically relevant variables based on previous experience are included for analysis. In contrast, machine learning is not built on a pre-structured model; rather, the data shape the model by detecting underlying patterns. The more variables (input) used to train the model, the more accurate the ultimate model will be. Leo Breiman distinguished two statistical modelling paradigms: the data model and the algorithmic model, wherein "algorithmic model" means more or less the machine learning algorithms like Random forest. Some statisticians have adopted methods from machine learning, producing the field of statistical learning. === Statistical physics === Analytical and computational techniques derived from deep-rooted physics of disordered systems can be extended to large-scale problems, including machine learning, e.g., to analyse the weight space of deep neural networks. Statistical physics is thus

  Read more →
• 

  Top 10 AI Analytics Tools Compared (2026)

  Shopping for the best AI analytics tool? An AI analytics tool is software that uses machine learning to help you get more done — it keeps getting smarter as the underlying models improve. Pricing, accuracy, and the size of the model behind the tool are the three factors that most affect daily usefulness. Whether you are a beginner or a pro, the right AI analytics tool slots into your workflow and pays for itself fast. Below we compare features, pricing, and real output so you can choose with confidence.

  Read more →
• 

  Pedro Domingos

  Pedro Domingos (born 1965) is a Professor Emeritus of computer science and engineering at the University of Washington. He is a researcher in machine learning known for Markov logic network enabling uncertain inference. == Education == Domingos received an undergraduate degree and Master of Science degree from Instituto Superior Técnico (IST). He moved to the University of California, Irvine, where he received a Master of Science degree followed by his PhD. == Research and career == After spending two years as an assistant professor at IST, he joined the University of Washington as an assistant professor of Computer Science and Engineering in 1999 and became a full professor in 2012. He started a machine learning research group at the hedge fund D. E. Shaw & Co. in 2018, but left in 2019. He co-founded the International Machine Learning Society. As of 2018, he was on the editorial board of Machine Learning journal. === Publications === Pedro Domingos, The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake Our World, New York, Basic Books, 2015, ISBN 978-0-465-06570-7. Pedro Domingos, "Our Digital Doubles: AI will serve our species, not control it", Scientific American, vol. 319, no. 3 (September 2018), pp. 88–93. "AIs are like autistic savants and will remain so for the foreseeable future.... AIs lack common sense and can easily make errors that a human never would... They are also liable to take our instructions too literally, giving us precisely what we asked for instead of what we actually wanted." (p. 93.) Pedro Domingos, 2040: A Silicon Valley Satire, BookBaby, 2024, ISBN 979-8-350-96334-2. === Awards and honors === 2014: ACM SIGKDD Innovation Award. for his foundational research in data stream analysis, cost-sensitive classification, adversarial learning, and Markov logic networks, as well as applications in viral marketing and information integration. 2010: Elected an Association for the Advancement of Artificial Intelligence (AAAI) Fellow. For significant contributions to the field of machine learning and to the unification of first-order logic and probability. 2003: Sloan Fellowship 1992–1997: Fulbright Scholarship

  Read more →
• 

  Keyword (linguistics)

  In corpus linguistics a key word is a word which occurs in a text more often than we would expect to occur by chance alone. Key words are calculated by carrying out a statistical test (e.g., loglinear or chi-squared) which compares the word frequencies in a text against their expected frequencies derived in a much larger corpus, which acts as a reference for general language use. Keyness is then the quality a word or phrase has of being "key" in its context. Combinations of nouns with parts of speech that human readers would not likely notice, such as prepositions, time adverbs, and pronouns can be a relevant part of keyness. Even separate pronouns can constitute keywords. Compare this with collocation, the quality linking two words or phrases usually assumed to be within a given span of each other. Keyness is a textual feature, not a language feature (so a word has keyness in a certain textual context but may well not have keyness in other contexts, whereas a node and collocate are often found together in texts of the same genre so collocation is to a considerable extent a language phenomenon). The set of keywords found in a given text share keyness, they are co-key. Words typically found in the same texts as a key word are called associates. == Sociological aspects == In politics, sociology and critical discourse analysis, the key reference for keywords was Raymond Williams (1976), but Williams was resolutely Marxist, and Critical Discourse Analysis has tended to perpetuate this political meaning of the term: keywords are part of ideologies and studying them is part of social criticism. Cultural studies has tended to develop along similar lines. This stands in stark contrast to present day linguistics which is wary of political analysis, and has tended to aspire to non-political objectivity. The development of technology, new techniques and methodology relating to massive corpora have all consolidated this trend. === Translatability === There are, however, numerous political dimensions that come into play when keywords are studied in relation to cultures, societies and their histories. The Lublin Ethnolinguistics School studies Polish and European keywords in this fashion. Anna Wierzbicka (1997), probably the best known cultural linguist writing in English today, studies languages as parts of cultures evolving in society and history. And it becomes impossible to ignore politics when keywords migrate from one culture to another. Underhill and Gianninoto demonstrate the way political terms like, "citizen" and "individual" are integrated into the Chinese worldview over the course of the 19th and 20th century. They argue that this is part of a complex readjustment of conceptual clusters related to "the people". Keywords like "citizen" generate various translations in Chinese, and are part of an ongoing adaptation to global concepts of individual rights and responsibilities. Understanding keywords in this light becomes crucial for understanding how the politics of China evolves as Communism emerges and as the free market and citizens' rights develop. Underhill and Gianninoto argue that this is part of the complex ways ideological worldviews interact with the language as an ongoing means of perceiving and understanding the world. Barbara Cassin studies keywords in a more traditional manner, striving to define the words specific to individual cultures, in order to demonstrate that many of our keywords are partially "untranslatable" into their "equivalents. The Greeks may need four words to cover all the meanings English-speakers have in mind when speaking of "love". Similarly, the French find that "liberté" suffices, while English-speakers attribute different associations to "liberty" and "freedom": "freedom of speech" or "freedom of movement", but "the Statue of Liberty". == Software-assisted identification == Keywords are identified by software that compares a word-list of the text with a word-list based on a larger reference corpus. Software such as e.g. WordSmith, lists keywords and phrases and allows plotting their occurrence as they appear in texts.

  Read more →
• 

  Fillrate

  In computer graphics, a video card's pixel fillrate refers to the number of pixels that can be rendered on the screen and written to video memory in one second. Pixel fillrates are given in megapixels per second or in gigapixels per second (in the case of newer cards), and are obtained by multiplying the number of render output units (ROPs) by the clock frequency of the graphics processing unit (GPU) of a video card. A similar concept, texture fillrate, refers to the number of texture map elements (texels) the GPU can map to pixels in one second. Texture fillrate is obtained by multiplying the number of texture mapping units (TMUs) by the clock frequency of the GPU. Texture fillrates are given in mega or gigatexels per second. However, there is no full agreement on how to calculate and report fillrates. Another possible method is to multiply the number of pixel pipelines by the GPU's clock frequency. The results of these multiplications correspond to a theoretical number. The actual fillrate depends on many other factors. In the past, the fillrate has been used as an indicator of performance by video card manufacturers such as ATI and NVIDIA, however, the importance of the fillrate as a measurement of performance has declined as the bottleneck in graphics applications has shifted. For example, today, the number and speed of unified shader processing units has gained attention. Although fillrate doesn't provide a substantial bottleneck in games, it can still provide a bottleneck for certain parts of the game, for example applying a gaussian blur can be bottlenecked by fillrate. Scene complexity can be increased by overdrawing, which happens when an object is drawn to the frame buffer, and another object (such as a wall) is then drawn on top of it, covering it up. The time spent drawing the first object is thus wasted because it is not visible. When a sequence of scenes is extremely complex (many pixels have to be drawn for each scene), the frame rate for the sequence may drop. When designing graphics intensive applications, one can determine whether the application is fillrate-limited (or shader limited) by seeing if the frame rate increases dramatically when the application runs at a lower resolution or in a smaller window. Although this is not a full-proof method, modern videogame engines can dynamically reduce the level-of-detail required and thereby reducing fillrate-limited applications. The best way to find fillrate bottlenecks is to use GPU vendor software like NVIDIA Nsight Graphics, AMD Radeon GPU Profile and the Intel Graphics Performance Analyzers.

  Read more →
• 

  Ofer Dekel (researcher)

  Ofer Dekel (Hebrew: עופר דקל) is a computer science researcher in the Machine Learning Department of Microsoft Research. He obtained his PhD in computer science from the Hebrew University of Jerusalem and is an affiliate faculty at the Computer Science & Engineering department at the University of Washington. == Areas of research == Dekel's research topics include machine learning, online prediction, statistical learning theory, and stochastic optimization. He is currently engaged in the application of machine learning techniques in the development of the Bing search engine.

  Read more →
• 

  Jaime Carbonell

  Jaime Guillermo Carbonell (July 29, 1953 – February 28, 2020) was a computer scientist who made seminal contributions to the development of natural language processing tools and technologies. His research in machine translation resulted in the development of several state-of-the-art language translation and artificial intelligence systems. He earned his B.S. degrees in Physics and in Mathematics from MIT in 1975 and did his Ph.D. under Dr. Roger Schank at Yale University in 1979. He joined Carnegie Mellon University as an assistant professor of computer science in 1979 and moved to Pittsburgh. He was affiliated with the Language Technologies Institute, Computer Science Department, Machine Learning Department, and Computational Biology Department at Carnegie Mellon. His interests spanned several areas of artificial intelligence, language technologies and machine learning. In particular, his research focused on areas such as text mining (extraction, categorization, novelty detection) and in new theoretical frameworks such as a unified utility-based theory bridging information retrieval, summarization, free-text question-answering and related tasks. He also worked on machine translation, both high-accuracy knowledge-based MT and machine learning for corpus-based MT (such as generalized example-based MT). == Career == Carbonell was the Allen Newell Professor of Computer Science and head of the Language Technologies Institute at Carnegie Mellon University. He joined Carnegie Mellon in 1979, and became a key faculty member in the artificial intelligence area. He was appointed full professor in 1987, Newell Chair in 1995, and University Professor in 2012. He completed his undergraduate studies at MIT. He received dual degrees in Mathematics and Physics. He received his Ph.D. in computer science from Yale University in 1979. At the time of his appointment, Carbonell was the youngest chaired professor in the School of Computer Science at CMU. His research spanned several areas of computer science, mostly in artificial intelligence, including: machine learning, data and text mining, natural language processing, very-large-scale knowledge bases, translingual information retrieval and automated summarization. He wrote more than 300 technical papers and gave over 500 invited or refereed-paper presentations (colloquia, seminars, panels, conferences, keynotes, etc.). He died following a long illness on February 28, 2020. Mona Talat Diab became the director of CMU's Language Technologies Institute in 2023. == Research == Carbonell created MMR (maximal marginal relevance) technology for text summarization and informational novelty detection in search engines, invention of transformational analogy, a generalized method for case-based reasoning (CBR) to re-use, modify and compose past successful plans for increasingly complex problems and knowledge-based interlingual machine translation. He was instrumental in setting up the Computational Biolinguistics Program, a joint venture between Carnegie Mellon and the University of Pittsburgh, which combines Language Technologies and Machine Learning to model and predict genomic, proteomic and glycomic 3D structures. Carbonell also did work in machine learning. He organized the first four machine learning conferences, starting with CMU in 1981. The Language Technologies Institute (LTI), founded and directed by Carbonell, achieved top honors in multiple areas. These areas include machine translation, search engines (including founding of Lycos by Michael Mauldin, one of Carbonell’s PhD students), speech synthesis, and education. LTI remains the original, largest and best-known institute for language technologies, with over $12M in annual funding and 200 researchers (faculty, staff, PhD students, MS students, visiting scholars etc.). Carbonell made major technical contributions in several fields, including (1) Creation of MMR (maximal marginal relevance) technology for text summarization and informational novelty detection in search engines,(2) Proactive machine learning for multi-source cost-sensitive active learning, (3) Linked conditional random fields for predicting tertiary and quaternary protein folds, (4) Symmetric optimal phrasal alignment method for trainable example-based and statistical machine translation, (5) Series- anomaly modeling for financial fraud detection and syndromic surveillance, (6) Knowledge-based interlingual machine translation, (7) Robust case-frame parsing, (8) Seeded version-space learning and (9) Invention of transformational and derivational analogy, generalized methods for case-based reasoning (CBR) to re-use, modify and compose past successful plans for increasingly complex problems. The teams led by Carbonell achieved top honors in many areas such as first scalable high-accuracy interlingual machine translation (1991), first speech-to-speech machine translation (1992), first large-scale spider and search engine (1994), and first trainable, large-scale protein-structure topology predictor (2005). Modern machine learning, co-founded by Carbonell, Michalski and Mitchell, is a fundamental enabling technology in search engines, data mining and social networking. Starting in 1980, he co-edited the first three books on ML, launched the ML conferences and was a co-founder and editor-in-chief of ML Journal. Carbonell’s innovations have led to several successful start-ups: Carnegie Group (AI expertsystems), Lycos (web search), Wisdom (financial optimization & ML), Carnegie Speech (spoken-language tutoring), Dynamix (data mining and pattern discovery), and Meaningful Machines (context-based machine translation). Carbonell was the founding director of The Language Technology Institute, the preeminent global institution in language studies, unparalleled in size and scope and has since been adopted/imitated in Germany (DFKI), Japan (Tokyo Univ.), and the US (Johns Hopkins). == Awards and honors == Okawa Prize, 2015 Best paper award, “Translingual Search” w/Yang, International Joint Conference on AI, 1997 Allen Newell endowed chair, Carnegie Mellon University, 1995 Elected fellow of AAAI, 1991 Computer Science teaching award, Carnegie Mellon University, 1987 Sperry Fellowship for excellence in AI research, 1986 Herbert Simon teaching award, 1986 "Recognition of Service" award from the ACM for the SIGART presidency, 1983–1985 Provided congressional testimony on machine translation, 1990 == Selected works == === Books === 1983. (with Ryszard S. Michalski & Tom M. Mitchell, Eds.) Machine learning: An artificial intelligence approach. Los Altos, CA: Morgan Kaufmann. 1986. (with Ryszard S. Michalski & Tom Mitchell, Eds.) Machine learning: An artificial intelligence approach. Vol. II. Los Altos, CA: Morgan-Kaufmann. 1986. (with Ryszard S. Michalski & Tom Mitchell, Eds.) Machine Learning: A Guide to Current Research. Kluwer Academic Publishers. == Contributions == “Protein Quaternary Fold Recognition Using Conditional Graphical Models” IJCAI 2007 (w/Liu et al.) “Context-Based Machine Translation” AMTA 2006 (w/Klein et al.) “SCRFs: A New Approach for Protein Fold Recognition,’’ Journal of Computational Biology, 13,2, 2006 (w/Liu et al) “MT for Resource-Poor Languages Using Elicitation-Based Learning” Machine Translation, 2004 ‘‘Learning Approaches for Detecting and Tracking News Events,’’ IEEE Trans I.S., 14, 4, 2000 (w/Yang)

  Read more →
• 

  Chelsea Finn

  Chelsea Finn (born October 8, 1992) is an American computer scientist and assistant professor at Stanford University. Her research investigates intelligence through the interactions of robots, with the hope to create robotic systems that can learn how to learn. She previously worked for Google and currently is a co-founder of the startup Physical Intelligence. == Early life and education == Finn was an undergraduate student in electrical engineering and computer science at Massachusetts Institute of Technology. She then moved to the University of California, Berkeley, where she earned her Ph.D. in 2018 under Pieter Abbeel and Sergey Levine. Her work in the Berkeley Artificial Intelligence Lab (BAIR) focused on gradient based algorithms . Such algorithms allow machines to 'learn to learn', more akin to human learning than traditional machine learning systems. These “meta-learning” techniques train machines to quickly adapt, such that when they encounter new scenarios they can learn quickly. As a doctoral student she worked as an intern at Google Brain, where she worked on robot learning algorithms from deep predictive models. She delivered a massive open online course on deep reinforcement learning. She was the first woman to win the C.V. & Daulat Ramamoorthy Distinguished Research Award. == Research and career == Finn investigates the capabilities of robots to develop intelligence through learning and interaction. She has made use of deep learning algorithms to simultaneously learn visual perception and control robotic skills. She developed meta-learning approaches to train neural networks to take in student code and output useful feedback. She showed that the system could quickly adapt without too much input from the instructor. She trialled the programme on Code in Place, a 12,000 student course delivered by Stanford University every year. She found that 97.9% of the time the students agreed with the feedback being given. == Awards and honors == 2016 C.V. & Daulat Ramamoorthy Distinguished Research Award 2017 Electrical engineering and computer science rising star 2018 MIT Technology Review 35 Under 35 2018 ACM Doctoral Dissertation Award 2020 Samsung Advanced Institute of Technology AI Researcher of the Year 2020 Intel Rising Star Faculty Award 2021 Office of Naval Research Young Investigator Award 2022 IEEE Robotics and Automation Society Early Academic Career Award == Select publications == Finn, Chelsea; Abbeel, Pieter; Levine, Sergey (2017-07-17). "Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks". International Conference on Machine Learning. PMLR: 1126–1135. arXiv:1703.03400. Sergey Levine; Chelsea Finn; Trevor Darrell; Pieter Abbeel (2016). "End-to-End Training of Deep Visuomotor Policies". Journal of Machine Learning Research. 17 (39): 1–40. arXiv:1504.00702. ISSN 1533-7928. Wikidata Q90313375. Chelsea Finn; Ian Goodfellow; Sergey Levine (2016). "Unsupervised Learning for Physical Interaction through Video Prediction" (PDF). Advances in Neural Information Processing Systems 29. Advances in Neural Information Processing Systems. Wikidata Q46993574.

  Read more →
• 

  Referring expression generation

  Referring expression generation (REG) is the subtask of natural language generation (NLG) that received most scholarly attention. While NLG is concerned with the conversion of non-linguistic information into natural language, REG focuses only on the creation of referring expressions (noun phrases) that identify specific entities called targets. This task can be split into two sections. The content selection part determines which set of properties distinguish the intended target and the linguistic realization part defines how these properties are translated into natural language. A variety of algorithms have been developed in the NLG community to generate different types of referring expressions. == Types of referring expressions == A referring expression (RE), in linguistics, is any noun phrase, or surrogate for a noun phrase, whose function in discourse is to identify some individual object (thing, being, event...) The technical terminology for identify differs a great deal from one school of linguistics to another. The most widespread term is probably refer, and a thing identified is a referent, as for example in the work of John Lyons. In linguistics, the study of reference relations belongs to pragmatics, the study of language use, though it is also a matter of great interest to philosophers, especially those wishing to understand the nature of knowledge, perception and cognition more generally. Various devices can be used for reference: determiners, pronouns, proper names... Reference relations can be of different kinds; referents can be in a "real" or imaginary world, in discourse itself, and they may be singular, plural, or collective. === Pronouns === The simplest type of referring expressions are pronoun such as he and it. The linguistics and natural language processing communities have developed various models for predicting anaphor referents, such as centering theory, and ideally referring-expression generation would be based on such models. However most NLG systems use much simpler algorithms, for example using a pronoun if the referent was mentioned in the previous sentence (or sentential clause), and no other entity of the same gender was mentioned in this sentence. === Definite noun phrases === There has been a considerable amount of research on generating definite noun phrases, such as the big red book. Much of this builds on the model proposed by Dale and Reiter. This has been extended in various ways, for example Krahmer et al. present a graph-theoretic model of definite NP generation with many nice properties. In recent years a shared-task event has compared different algorithms for definite NP generation, using the TUNA corpus. === Spatial and temporal reference === Recently there has been more research on generating referring expressions for time and space. Such references tend to be imprecise (what is the exact meaning of tonight?), and also to be interpreted in different ways by different people. Hence it may be necessary to explicitly reason about false positive vs false negative tradeoffs, and even calculate the utility of different possible referring expressions in a particular task context. === Criteria for good expressions === Ideally, a good referring expression should satisfy a number of criteria: Referential success: It should unambiguously identify the referent to the reader. Ease of comprehension: The reader should be able to quickly read and understand it. Computational complexity: The generation algorithm should be fast No false inferences: The expression should not confuse or mislead the reader by suggesting false implicatures or other pragmatic inferences. For example, a reader may be confused if he is told Sit by the brown wooden table in a context where there is only one table. == History == === Pre-2000 era === REG goes back to the early days of NLG. One of the first approaches was done by Winograd in 1972 who developed an "incremental" REG algorithm for his SHRDLU program. Afterwards researchers started to model the human abilities to create referring expressions in the 1980s. This new approach to the topic was influenced by the researchers Appelt and Kronfeld who created the programs KAMP and BERTRAND and considered referring expressions as parts of bigger speech acts. Some of their most interesting findings were the fact that referring expressions can be used to add information beyond the identification of the referent as well as the influence of communicative context and the Gricean maxims on referring expressions. Furthermore, its skepticism concerning the naturalness of minimal descriptions made Appelt and Kronfeld's research a foundation of later work on REG. The search for simple, well-defined problems changed the direction of research in the early 1990s. This new approach was led by Dale and Reiter who stressed the identification of the referent as the central goal. Like Appelt they discuss the connection between the Gricean maxims and referring expressions in their culminant paper in which they also propose a formal problem definition. Furthermore, Reiter and Dale discuss the Full Brevity and Greedy Heuristics algorithms as well as their Incremental Algorithm(IA) which became one of the most important algorithms in REG. === Later developments === After 2000 the research began to lift some of the simplifying assumptions, that had been made in early REG research in order to create more simple algorithms. Different research groups concentrated on different limitations creating several expanded algorithms. Often these extend the IA in a single perspective for example in relation to: Reference to Sets like "the t-shirt wearers" or "the green apples and the banana on the left" Relational Descriptions like "the cup on the table" or "the woman who has three children" Context Dependency, Vagueness and Gradeability include statements like "the older man" or "the car on the left" which are often unclear without a context Salience and Generation of Pronouns are highly discourse dependent making for example "she" a reference to "the (most salient) female person" Many simplifying assumptions are still in place or have just begun to be worked on. Also a combination of the different extensions has yet to be done and is called a "non-trivial enterprise" by Krahmer and van Deemter. Another important change after 2000 was the increasing use of empirical studies in order to evaluate algorithms. This development took place due to the emergence of transparent corpora. Although there are still discussions about what the best evaluation metrics are, the use of experimental evaluation has already led to a better comparability of algorithms, a discussion about the goals of REG and more task-oriented research. Furthermore, research has extended its range to related topics such as the choice of Knowledge Representation(KR) Frameworks. In this area the main question, which KR framework is most suitable for the use in REG remains open. The answer to this question depends on how well descriptions can be expressed or found. A lot of the potential of KR frameworks has been left unused so far. Some of the different approaches are the usage of: Graph search which treats relations between targets in the same way as properties. Constraint Satisfaction which allows for a separation between problem specification and the implementation. Modern Knowledge Representation which offers logical inference in for example Description Logic or Conceptual Graphs. == Problem definition == Dale and Reiter (1995) think about referring expressions as distinguishing descriptions. They define: The referent as the entity that should be described The context set as set of salient entities The contrast set or potential distractors as all elements of the context set except the referent A property as a reference to a single attribute–value pair Each entity in the domain can be characterised as a set of attribute–value pairs for example ⟨ {\displaystyle \langle } type, dog ⟩ {\displaystyle \rangle } , ⟨ {\displaystyle \langle } gender, female ⟩ {\displaystyle \rangle } or ⟨ {\displaystyle \langle } age, 10 years ⟩ {\displaystyle \rangle } . The problem then is defined as follows: Let r {\displaystyle r} be the intended referent, and C {\displaystyle C} be the contrast set. Then, a set L {\displaystyle L} of attribute–value pairs will represent a distinguishing description if the following two conditions hold: Every attribute–value pair in L {\displaystyle L} applies to r {\displaystyle r} : that is, every element of L {\displaystyle L} specifies an attribute–value that r {\displaystyle r} possesses. For every member c {\displaystyle c} of C {\displaystyle C} , there is at least one element l {\displaystyle l} of L {\displaystyle L} that does not apply to c {\displaystyle c} : that is, there is an l {\displaystyle l} in L {\displaystyle L} that specifies an attribute–value that c {\displaystyle c} does not possess. l {\displaystyle l} is said

  Read more →
• 

  Amazon Polly

  Amazon Polly is a cloud service by Amazon Web Services, a subsidiary of Amazon.com, that converts text into spoken audio. It allows developers to create speech-enabled applications and products. It was launched in November 2016 and (as of December 2024) includes 100+ voices across 41 language variants, some of which are Neural Text-to-Speech voices of higher quality. Users include Duolingo, a language education platform.

  Read more →
• 

  Unsupervised learning

  Unsupervised learning is a framework in machine learning where, in contrast to supervised learning, algorithms learn patterns exclusively from unlabeled data. Other frameworks in the spectrum of supervisions include weak- or semi-supervision, where a small portion of the data is tagged, and self-supervision. Some researchers consider self-supervised learning a form of unsupervised learning. Conceptually, unsupervised learning divides into the aspects of data, training, algorithm, and downstream applications. Typically, the dataset is harvested cheaply "in the wild", such as massive text corpus obtained by web crawling, with only minor filtering (such as Common Crawl). This compares favorably to supervised learning, where the dataset (such as the ImageNet1000) is typically constructed manually, which is much more expensive. There are algorithms designed specifically for unsupervised learning, such as clustering algorithms like k-means, dimensionality reduction techniques like principal component analysis (PCA), Boltzmann machine learning, and autoencoders. After the rise of deep learning, most large-scale unsupervised learning has been done by training general-purpose neural network architectures by gradient descent, adapted to performing unsupervised learning by designing an appropriate training procedure. Sometimes a trained model can be used as-is, but more often they are modified for downstream applications. For example, the generative pretraining method trains a model to generate a textual dataset, before finetuning it for other applications, such as text classification. As another example, autoencoders are trained to produce good features, which can then be used as a module for other models, such as in a latent diffusion model. == Tasks == Tasks are often categorized as discriminative (recognition) or generative (imagination). Often but not always, discriminative tasks use supervised methods and generative tasks use unsupervised (see Venn diagram); however, the separation is very hazy. For example, object recognition favors supervised learning but unsupervised learning can also cluster objects into groups. Furthermore, as progress marches onward, some tasks employ both methods, and some tasks swing from one to another. For example, image recognition started off as heavily supervised, but became hybrid by employing unsupervised pre-training, and then moved towards supervision again with the advent of dropout, ReLU, and adaptive learning rates. A typical generative task is as follows. At each step, a datapoint is sampled from the dataset, and part of the data is removed, and the model must infer the removed part. This is particularly clear for the denoising autoencoders and BERT. == Neural network architectures == === Training === During the learning phase, an unsupervised network tries to mimic the data it is given and uses the error in its mimicked output to correct itself (i.e. correct its weights and biases). Sometimes the error is expressed as a low probability that the erroneous output occurs, or it might be expressed as an unstable high energy state in the network. In contrast to supervised methods' dominant use of backpropagation, unsupervised learning also employs other methods including: Hopfield learning rule, Boltzmann learning rule, Contrastive Divergence, Wake Sleep, Variational Inference, Maximum Likelihood, Maximum A Posteriori, Gibbs Sampling, and backpropagating reconstruction errors or hidden state reparameterizations. See the table below for more details. === Energy === An energy function is a macroscopic measure of a network's activation state. In Boltzmann machines, it plays the role of the Cost function. This analogy with physics is inspired by Ludwig Boltzmann's analysis of a gas' macroscopic energy from the microscopic probabilities of particle motion p ∝ e − E / k T {\displaystyle p\propto e^{-E/kT}} , where k is the Boltzmann constant and T is temperature. In the RBM network the relation is p = e − E / Z {\displaystyle p=e^{-E}/Z} , where p {\displaystyle p} and E {\displaystyle E} vary over every possible activation pattern and Z = ∑ All Patterns e − E ( pattern ) {\displaystyle \textstyle {Z=\sum _{\scriptscriptstyle {\text{All Patterns}}}e^{-E({\text{pattern}})}}} . To be more precise, p ( a ) = e − E ( a ) / Z {\displaystyle p(a)=e^{-E(a)}/Z} , where a {\displaystyle a} is an activation pattern of all neurons (visible and hidden). Hence, some early neural networks bear the name Boltzmann Machine. Paul Smolensky calls − E {\displaystyle -E\,} the Harmony. A network seeks low energy which is high Harmony. === Networks === This table shows connection diagrams of various unsupervised networks, the details of which will be given in the section Comparison of Networks. Circles are neurons and edges between them are connection weights. As network design changes, features are added on to enable new capabilities or removed to make learning faster. For instance, neurons change between deterministic (Hopfield) and stochastic (Boltzmann) to allow robust output, weights are removed within a layer (RBM) to hasten learning, or connections are allowed to become asymmetric (Helmholtz). Of the networks bearing people's names, only Hopfield worked directly with neural networks. Boltzmann and Helmholtz came before artificial neural networks, but their work in physics and physiology inspired the analytical methods that were used. === History === === Specific Networks === Here, we highlight some characteristics of select networks. The details of each are given in the comparison table below. Hopfield Network Ferromagnetism inspired Hopfield networks. A neuron corresponds to an iron domain with binary magnetic moments Up and Down, and neural connections correspond to the domain's influence on each other. Symmetric connections enable a global energy formulation. During inference the network updates each state using the standard activation step function. Symmetric weights and the right energy functions guarantees convergence to a stable activation pattern. Asymmetric weights are difficult to analyze. Hopfield nets are used as Content Addressable Memories (CAM). Boltzmann Machine These are stochastic Hopfield nets. Their state value is sampled from this pdf as follows: suppose a binary neuron fires with the Bernoulli probability p(1) = 1/3 and rests with p(0) = 2/3. One samples from it by taking a uniformly distributed random number y, and plugging it into the inverted cumulative distribution function, which in this case is the step function thresholded at 2/3. The inverse function = { 0 if x <= 2/3, 1 if x > 2/3 }. Sigmoid Belief Net Introduced by Radford Neal in 1992, this network applies ideas from probabilistic graphical models to neural networks. A key difference is that nodes in graphical models have pre-assigned meanings, whereas Belief Net neurons' features are determined after training. The network is a sparsely connected directed acyclic graph composed of binary stochastic neurons. The learning rule comes from Maximum Likelihood on p(X): Δwij ∝ {\displaystyle \propto } sj (si - pi), where pi = 1 / ( 1 + eweighted inputs into neuron i ). sj's are activations from an unbiased sample of the posterior distribution and this is problematic due to the Explaining Away problem raised by Judea Perl. Variational Bayesian methods uses a surrogate posterior and blatantly disregard this complexity. Deep Belief Network Introduced by Hinton, this network is a hybrid of RBM and Sigmoid Belief Network. The top 2 layers is an RBM and the second layer downwards form a sigmoid belief network. One trains it by the stacked RBM method and then throw away the recognition weights below the top RBM. As of 2009, 3-4 layers seems to be the optimal depth. Helmholtz machine These are early inspirations for the Variational Auto Encoders. Its 2 networks combined into one—forward weights operates recognition and backward weights implements imagination. It is perhaps the first network to do both. Helmholtz did not work in machine learning but he inspired the view of "statistical inference engine whose function is to infer probable causes of sensory input". the stochastic binary neuron outputs a probability that its state is 0 or 1. The data input is normally not considered a layer, but in the Helmholtz machine generation mode, the data layer receives input from the middle layer and has separate weights for this purpose, so it is considered a layer. Hence this network has 3 layers. Variational autoencoder These are inspired by Helmholtz machines and combines probability network with neural networks. An Autoencoder is a 3-layer CAM network, where the middle layer is supposed to be some internal representation of input patterns. The encoder neural network is a probability distribution qφ(z given x) and the decoder network is pθ(x given z). The weights are named phi & theta rather than W and V as in Helmholtz—a cosmetic difference. These 2 networks h

  Read more →