Blocknots were random sequences of numbers contained in a book and organized by numbered rows and columns and were used as additives in the reciphering of Soviet Union codes, during World War II. The Blocknot consisted of a booklet of fifty sheets of 5-figure random additive, 100 additive groups to a sheet. No sheet was used more than once, thus the blocknots were in effect a form of one-time pad. The Soviet Unions highest grade ciphers that were used in the East, were the 5-figure codebook enciphered with the Blocknot book, and were generally considered unbreakable. == Technical Description == Blocknots were distributed centrally from an office in Moscow. Every Blocknot contained 5-figure groups in a number of sheets, for the enciphering of 5-figure messages. The encipherment was effected by applying additives taken from the pad, of which 50-100 5-figure groups appeared. Each pad had a 5-figure number and each sheet had a 2-figure number running consecutively. There were 5 different types of Blocknots, in two different categories The Individual in which each table of random numbers was used only once. The General in which each page of the Blocknot was valid for one day. The security of the additive sequence rested on the choice of different starting points for each message. In 5-figure messages, the blocknot was one of the first 10 Groups in the message. Its position changed at long intervals, but was always easy to re-identify. The Russians differentiated between three types of blocks: The 3-block, DRIERBLOCK. I-block for Individual Block: 50 pages, additive read off in one direction only. The messages could be used and read only between 2 wireless telegraphy stations on one net. The 6-block, SECHSERBLOCK. Z-block for Circular Block: 30 pages, additive read off in either direction. The messages could be used and read, between all W/T stations in a net. The 2-block, ZWEIERBLOCK. OS-block. Used only in traffic from lower to higher formations. Two other types were used, in lower echelons. Notblock: Used in an emergency. Blocknot used for passing on traffic. The distribution of Blocknots was carried out centrally from Moscow to Army Groups then to Armies. The Army was responsible for their distribution throughout the lower levels of the army down to company level. Independent units took their cipher material with them. Occasionally the same blocknot was distributed to two units on different parts of the front, which enabled Depth to be established. Records of all Blocknots used were kept in Berlin and when a repeat was noticed a BLOCKNOT ANGEBOT message was sent out to all German Signals units, to indicate that it may have been possible to break the code using it. There was no certainty in this. A cryptanalyst with the General der Nachrichtenaufklärung stated while being interrogated by TICOM: It seems that depths of up to 8 were established at the beginning of the Russian Campaign but that no 5-figure code was broken after May 1943 German cryptanalysts who were prisoners of war stated under interrogation, that each of the figures 0 to 9 were placed en clair usually within the first ten groups of the text or sometimes at the end. One indicator was the Blocknot number and the consisted of two random figures, the figure representing the type, and the remaining two, the page of the Blocknot being used. In long messages, 000000 was placed in the message when the end of a page had been reached. == Chi number == The Chi-number was the serial numbering of all 5-figure messages passing through the hands of the Cipher Officer, starting on the first of January and ending on thirty-first December of the current year. It always appeared as the last group in an intercepted message, e.g. 00001 on the 1st January, or when the unit was newly set up. The progression of Chi-numbers was carefully observed and recorded in the form of a graph. A Russian corps had about 10 5-figure messages per day, and Army about 20-30 and a Front about 60–100. After only a relatively short time, the individual curves separated sharply and the type of formation could be recognized by the height of the Chi-number alone. == Monitoring == Blocknots were tracked in a card index, that was maintained by the Signal Intelligence Evaluation Centre (NAAS). The NAAS functionality included evaluation and traffic analysis, cryptanalysis, collation and dissemination of intelligence. The card index, which was one amongst several Card Indexes. A careful recording and study of blocks provided the positive clues in the identification and tracking of formations using 5-figure ciphers. The index was subdivided into two files: Search card index, contained all blocknots and chi-numbers whether or not they were known. Unit card index, contained only known Block and Chi-numbers. Inspector Berger, who was the chief cryptanalyst of NAAS 1 stated that the two files formed: The most important and surest instruments for identifying Russian radio nets, known to him. The Blocknots were also used in the Stationary Intercept Company (Feste), the military unit that were designed to work at a lower level to the NAAS, at the Army level and were semi-motorized, and closer to the front. The Feste used the Blocknot value along with several other parameters to build a network diagram. The network diagram was studied extensively, as part of a 6-stage process, that involved several departments within the Feste. The outcome was a metric which determined the most interesting circuit for traffic monitoring, and least interesting, where monitoring of traffic should cease. == Analysis == Johannes Marquart was a mathematician and cryptanalyst who initially worked for Inspectorate 7/VI and later led Referat Ia of Group IV of the General der Nachrichtenaufklärung. Marquart was assigned the study of the Soviet Union Blocknot traffic. Marquart and his unit conducted extensive research in an attempt to discover the method by which they were produced. All the counts which they made, however, failed to reveal any non-random characteristics in the design of the tables, and while they thought the Blocknots must have been generated by machine, they were never able to draw any concrete deductions as a result of their research. == Example == The Soviet 3rd Guard Tank Army transmits a 5-figure message with the Blocknot of 37581 (one of the first 10 groups in the message). On the same day the Block 37582 was used by the same formation. The next day 37583 appeared. Thereafter, for a period, the Army was not heard by German Wireless telegraphy intercept operators, as it was maintaining wireless silence. After a few days, an unidentified net with the Blocknot 37588 is picked up. This message net is claimed, because of the proximity of the blocks (88/83) to be the 3rd Guard Tank Army. The missing Blocknots 84-87 were presumably used in telegraphic, telephonic or courier communications. The Chi number provides confirmation of the first assumption, based on proximity of blocknots in most cases.
Random feature
Random features (RF) are a technique used in machine learning to approximate kernel methods, introduced by Ali Rahimi and Ben Recht in their 2007 paper "Random Features for Large-Scale Kernel Machines", and extended by. RF uses a Monte Carlo approximation to kernel functions by randomly sampled feature maps. It is used for datasets that are too large for traditional kernel methods like support vector machine, kernel ridge regression, and gaussian process. == Mathematics == === Kernel method === Given a feature map ϕ : R d → V {\textstyle \phi :\mathbb {R} ^{d}\to V} , where V {\textstyle V} is a Hilbert space (more specifically, a reproducing kernel Hilbert space), the kernel trick replaces inner products in feature space ⟨ ϕ ( x i ) , ϕ ( x j ) ⟩ V {\displaystyle \langle \phi (x_{i}),\phi (x_{j})\rangle _{V}} by a kernel function k ( x i , x j ) : R d × R d → R {\displaystyle k(x_{i},x_{j}):\mathbb {R} ^{d}\times \mathbb {R} ^{d}\to \mathbb {R} } Kernel methods replaces linear operations in high-dimensional space by operations on the kernel matrix: K X := [ k ( x i , x j ) ] i , j ∈ 1 : N {\displaystyle K_{X}:=[k(x_{i},x_{j})]_{i,j\in 1:N}} where N {\textstyle N} is the number of data points. === Random kernel method === The problem with kernel methods is that the kernel matrix K X {\textstyle K_{X}} has size N × N {\textstyle N\times N} . This becomes computationally infeasible when N {\textstyle N} reaches the order of a million. The random kernel method replaces the kernel function k {\textstyle k} by an inner product in low-dimensional feature space R D {\textstyle \mathbb {R} ^{D}} : k ( x , y ) ≈ ⟨ z ( x ) , z ( y ) ⟩ {\displaystyle k(x,y)\approx \langle z(x),z(y)\rangle } where z {\textstyle z} is a randomly sampled feature map z : R d → R D {\textstyle z:\mathbb {R} ^{d}\to \mathbb {R} ^{D}} . This converts kernel linear regression into linear regression in feature space, kernel SVM into SVM in feature space, etc. Since we have K X ≈ Z X T Z X {\displaystyle K_{X}\approx Z_{X}^{T}Z_{X}} where Z X = [ z ( x 1 ) , … , z ( x N ) ] {\displaystyle Z_{X}=[z(x_{1}),\dots ,z(x_{N})]} , these methods no longer involve matrices of size O ( N 2 ) {\textstyle O(N^{2})} , but only random feature matrices of size O ( D N ) {\textstyle O(DN)} . == Random Fourier feature == === Radial basis function kernel === The radial basis function (RBF) kernel on two samples x i , x j ∈ R d {\displaystyle x_{i},x_{j}\in \mathbb {R} ^{d}} is defined as k ( x i , x j ) = exp ( − ‖ x i − x j ‖ 2 2 σ 2 ) {\displaystyle k(x_{i},x_{j})=\exp \left(-{\frac {\|x_{i}-x_{j}\|^{2}}{2\sigma ^{2}}}\right)} where ‖ x i − x j ‖ 2 {\displaystyle \|x_{i}-x_{j}\|^{2}} is the squared Euclidean distance and σ {\displaystyle \sigma } is a free parameter defining the shape of the kernel. It can be approximated by a random Fourier feature map z : R d → R 2 D {\displaystyle z:\mathbb {R} ^{d}\to \mathbb {R} ^{2D}} : z ( x ) := 1 D [ cos ⟨ ω 1 , x ⟩ , sin ⟨ ω 1 , x ⟩ , … , cos ⟨ ω D , x ⟩ , sin ⟨ ω D , x ⟩ ] T {\displaystyle z(x):={\frac {1}{\sqrt {D}}}[\cos \langle \omega _{1},x\rangle ,\sin \langle \omega _{1},x\rangle ,\ldots ,\cos \langle \omega _{D},x\rangle ,\sin \langle \omega _{D},x\rangle ]^{T}} where ω 1 , . . . , ω D {\displaystyle \omega _{1},...,\omega _{D}} are IID samples from the multidimensional normal distribution N ( 0 , σ − 2 I ) {\displaystyle N(0,\sigma ^{-2}I)} . Since cos , sin {\displaystyle \cos ,\sin } are bounded, there is a stronger convergence guarantee by Hoeffding's inequality. === Random Fourier features === By Bochner's theorem, the above construction can be generalized to arbitrary positive definite shift-invariant kernel k ( x , y ) = k ( x − y ) {\displaystyle k(x,y)=k(x-y)} . Define its Fourier transform p ( ω ) = 1 2 π ∫ R d e − j ⟨ ω , Δ ⟩ k ( Δ ) d Δ {\displaystyle p(\omega )={\frac {1}{2\pi }}\int _{\mathbb {R} ^{d}}e^{-j\langle \omega ,\Delta \rangle }k(\Delta )d\Delta } then ω 1 , . . . , ω D {\displaystyle \omega _{1},...,\omega _{D}} are sampled IID from the probability distribution with probability density p {\displaystyle p} . This applies for other kernels like the Laplace kernel and the Cauchy kernel. === Neural network interpretation === Given a random Fourier feature map z {\displaystyle z} , training the feature on a dataset by featurized linear regression is equivalent to fitting complex parameters θ 1 , … , θ D ∈ C {\displaystyle \theta _{1},\dots ,\theta _{D}\in \mathbb {C} } such that f θ ( x ) = R e ( ∑ k θ k e i ⟨ ω k , x ⟩ ) {\displaystyle f_{\theta }(x)=\mathrm {Re} \left(\sum _{k}\theta _{k}e^{i\langle \omega _{k},x\rangle }\right)} which is a neural network with a single hidden layer, with activation function t ↦ e i t {\displaystyle t\mapsto e^{it}} , zero bias, and the parameters in the first layer frozen. In the overparameterized case, when 2 D ≥ N {\displaystyle 2D\geq N} , the network linearly interpolates the dataset { ( x i , y i ) } i ∈ 1 : N {\displaystyle \{(x_{i},y_{i})\}_{i\in 1:N}} , and the network parameters is the least-norm solution: θ ^ = arg min θ ∈ C D , f θ ( x k ) = y k ∀ k ∈ 1 : N ‖ θ ‖ {\displaystyle {\hat {\theta }}=\arg \min _{\theta \in \mathbb {C} ^{D},f_{\theta }(x_{k})=y_{k}\forall k\in 1:N}\|\theta \|} At the limit of D → ∞ {\displaystyle D\to \infty } , the L2 norm ‖ θ ^ ‖ → ‖ f K ‖ H {\displaystyle \|{\hat {\theta }}\|\to \|f_{K}\|_{H}} where f K {\displaystyle f_{K}} is the interpolating function obtained by the kernel regression with the original kernel, and ‖ ⋅ ‖ H {\displaystyle \|\cdot \|_{H}} is the norm in the reproducing kernel Hilbert space for the kernel. == Other examples == === Random binning features === A random binning features map partitions the input space using randomly shifted grids at randomly chosen resolutions and assigns to an input point a binary bit string that corresponds to the bins in which it falls. The grids are constructed so that the probability that two points x i , x j ∈ R d {\displaystyle x_{i},x_{j}\in \mathbb {R} ^{d}} are assigned to the same bin is proportional to K ( x i , x j ) {\displaystyle K(x_{i},x_{j})} . The inner product between a pair of transformed points is proportional to the number of times the two points are binned together, and is therefore an unbiased estimate of K ( x i , x j ) {\displaystyle K(x_{i},x_{j})} . Since this mapping is not smooth and uses the proximity between input points, Random Binning Features works well for approximating kernels that depend only on the L 1 {\displaystyle L_{1}} distance between datapoints. === Orthogonal random features === Orthogonal random features uses a random orthogonal matrix instead of a random Fourier matrix. == Historical context == In NIPS 2006, deep learning had just become competitive with linear models like PCA and linear SVMs for large datasets, and people speculated about whether it could compete with kernel SVMs. However, there was no way to train kernel SVM on large datasets. The two authors developed the random feature method to train those. It was then found that the O ( 1 / D ) {\displaystyle O(1/D)} variance bound did not match practice: the variance bound predicts that approximation to within 0.01 {\displaystyle 0.01} requires D ∼ 10 4 {\displaystyle D\sim 10^{4}} , but in practice required only ∼ 10 2 {\displaystyle \sim 10^{2}} . Attempting to discover what caused this led to the subsequent two papers.
Distributed artificial intelligence
Distributed Artificial Intelligence (DAI) (also called Decentralized Artificial Intelligence) is a melding of artificial intelligence with distributed computing. From artificial intelligence comes the theory and technology for constructing or analyzing an intelligent system. But where artificial intelligence uses psychology as a source of ideas, inspiration, and metaphor, DAI uses sociology, economics, and management science for inspiration. Where the focus of artificial intelligence is on the individual, the focus of DAI is on the group. Distributed computing provides the computational substrate on which this group focus can occur. Using techniques from artificial intelligence, communication theory, control theory, and interaction theory, it produces a cooperative solution to problems by a decentralized group of computational entities (agents). DAI is closely related to and a predecessor of the field of multi-agent systems. They are distinguished generally by multi-agent systems being open, where the entities might arise from different interests and have individual goals, and distributed artificial-intelligence systems, where the entities have common goals. There are numerous applications and tools. == Definition == Distributed Artificial Intelligence (DAI) is an approach to solving complex learning, planning, and decision-making problems. It is embarrassingly parallel, thus able to exploit large scale computation and spatial distribution of computing resources. These properties allow it to solve problems that require the processing of very large data sets. DAI systems consist of autonomous learning processing nodes (agents), that are distributed, often at a very large scale. DAI nodes can act independently, and partial solutions are integrated by communication between nodes, often asynchronously. By virtue of their scale, DAI systems are robust and elastic, and by necessity, loosely coupled. Furthermore, DAI systems are built to be adaptive to changes in the problem definition or underlying data sets due to the scale and difficulty in redeployment. DAI systems do not require all the relevant data to be aggregated in a single location, in contrast to monolithic or centralized Artificial Intelligence systems, which have tightly coupled and geographically close processing nodes. Therefore, DAI systems often operate on sub-samples or hashed impressions of very large datasets. In addition, the source dataset may change or be updated during the course of the execution of a DAI system. == Development == In 1975 distributed artificial intelligence emerged as a subfield of artificial intelligence that dealt with interactions of intelligent agents. As a scientific discipline, it progressed through a series of workshops in the USA (International Workshop on Distributed Artificial Intelligence, held in 13 editions from 1978 - 1994), Europe (Workshop on Modelling Autonomous Agents in a Multi-Agent World https://link.springer.com/conference/maamaw), and Asia (Multi-Agent and Cooperative Computation Workshop (MACC) https://sites.google.com/view/sig-macc/macc-workshop?authuser=0). Distributed artificial intelligence systems were conceived as a group of intelligent entities, called agents, that interacted by cooperation, by coexistence, or by competition. DAI is categorized into multi-agent systems and distributed problem solving. In multi-agent systems the main focus is how agents coordinate their knowledge and activities. For distributed problem solving the major focus is how the problem is decomposed and the solutions are synthesized. == Goals == The objectives of Distributed Artificial Intelligence are to solve the reasoning, planning, learning and perception problems of artificial intelligence, especially if they require large data, by distributing the problem to autonomous processing nodes (agents). To reach the objective, DAI requires: A distributed system with robust and elastic computation on unreliable and failing resources that are loosely coupled Coordination of the actions and communication of the nodes Subsamples of large data sets and online machine learning There are many reasons for wanting to distribute intelligence or cope with multi-agent systems. Mainstream problems in DAI research include the following: Parallel problem solving: mainly deals with how classic artificial intelligence concepts can be modified, so that multiprocessor systems and clusters of computers can be used to speed up calculation. Distributed problem solving (DPS): the concept of agent, autonomous entities that can communicate with each other, was developed to serve as an abstraction for developing DPS systems. See below for further details. Multi-Agent Based Simulation (MABS): a branch of DAI that builds the foundation for simulations that need to analyze not only phenomena at macro level but also at micro level, as it is in many social simulation scenarios. == Approaches == Two types of DAI has emerged: In Multi-agent systems agents coordinate their knowledge and activities and reason about the processes of coordination. Agents are physical or virtual entities that can act, perceive their environment, and communicate with other agents. An agent is autonomous and has skills to achieve goals. The agents change the state of their environment by their actions. There are a number of different coordination techniques. In distributed problem solving the work is divided among nodes and the knowledge is shared. The main concerns are task decomposition and synthesis of the knowledge and solutions. DAI can apply a bottom-up approach to AI, similar to the subsumption architecture as well as the traditional top-down approach of AI. In addition, DAI can also be a vehicle for emergence. === Challenges === The challenges in Distributed AI are: How to carry out communication and interaction of agents and which communication language or protocols should be used. How to ensure the coherency of agents. How to synthesise the results among 'intelligent agents' group by formulation, description, decomposition and allocation. == Applications and tools == Areas where DAI have been applied are: Electronic commerce, e.g. for trading strategies the DAI system learns financial trading rules from subsamples of very large samples of financial data Networks, e.g. in telecommunications the DAI system controls the cooperative resources in a WLAN network Routing, e.g. model vehicle flow in transport networks Scheduling, e.g. flow shop scheduling where the resource management entity ensures local optimization and cooperation for global and local consistency Search engines, e.g. in LLM federated search like Ithy where document retrieval and analysis are distributed to DAI agents before aggregation Multi-Agent systems, e.g. artificial life, the study of simulated life Electric power systems, e.g. Condition Monitoring Multi-Agent System (COMMAS) applied to transformer condition monitoring, and IntelliTEAM II Automatic Restoration System DAI integration in tools has included: ECStar is a distributed rule-based learning system. == Agents == === Systems: Agents and multi-agents === Notion of Agents: Agents can be described as distinct entities with standard boundaries and interfaces designed for problem solving. Notion of Multi-Agents: Multi-Agent system is defined as a network of agents which are loosely coupled working as a single entity like society for problem solving that an individual agent cannot solve. === Software agents === The key concept used in DPS and MABS is the abstraction called software agents. An agent is a virtual (or physical) autonomous entity that has an understanding of its environment and acts upon it. An agent is usually able to communicate with other agents in the same system to achieve a common goal, that one agent alone could not achieve. This communication system uses an agent communication language. A first classification that is useful is to divide agents into: reactive agent – A reactive agent is not much more than an automaton that receives input, processes it and produces an output. deliberative agent – A deliberative agent in contrast should have an internal view of its environment and is able to follow its own plans. hybrid agent – A hybrid agent is a mixture of reactive and deliberative, that follows its own plans, but also sometimes directly reacts to external events without deliberation. Well-recognized agent architectures that describe how an agent is internally structured are: ASMO (emergence of distributed modules) BDI (Believe Desire Intention, a general architecture that describes how plans are made) InterRAP (A three-layer architecture, with a reactive, a deliberative and a social layer) PECS (Physics, Emotion, Cognition, Social, describes how those four parts influences the agents behavior). Soar (a rule-based approach)
Xaitment
xaitment is a German-based company that develops and sells artificial intelligence (AI) software to video game developers and simulation developers. The company was founded in 2004 by Dr. Andreas Gerber, and is a spin-off of the German Research Centre for Artificial Intelligence, or DFKI. xaitment has its main office in Quierschied, Germany, and field offices in San Francisco and China. == Products == xaitment currently sells two AI software modules: xaitMap and xaitControl. xaitMap provides runtime libraries and graphical tools for navigation mesh generation (also called NavMesh generation), pathfinding, dynamic collision avoidance, and individual and crowd movement. xaitControl is a finite-state machine for game logic and character behavior modeling that also includes a real-time debugger. On January 11, 2012, xaitment announced that it making its source code for these modules available to "all current and future US and European licensees". On February 22, 2012 xaitment released two new plug-ins, xaitMap and xaitControl for the Unity Game Engine. The full versions are available for PC (Windows and Linux), PlayStation 3, Xbox 360 and Wii. The pathfinding plug-in is available with a Windows dev environment, but can deployed on iOS, Mac, Android and the Unity Web Player. == Partners == xaitment's AI software is currently integrated into the Unity game engine, Havok's Vision Engine, Bohemia Interactive's VBS2 Simulation Engine, GameBase's Gamebryo game engine. == Customers == xaitment sells its AI software products to video game developers and military and civil simulation developers. Current customers include Tencent, gamania, TML Studios, Emobi Games, IP Keys and others. A full list of customers can be found on xaitment's website.
Speech-generating device
Speech-generating devices (SGDs), also known as voice output communication aids, are electronic augmentative and alternative communication (AAC) systems used to supplement or replace speech or writing for individuals with severe speech impairments, enabling them to verbally communicate. SGDs are important for people who have limited means of interacting verbally, as they allow individuals to become active participants in communication interactions. They are particularly helpful for patients with amyotrophic lateral sclerosis (ALS) but recently have been used for children with predicted speech deficiencies. There are several input and display methods for users of varying abilities to make use of SGDs. Some SGDs have multiple pages of symbols to accommodate a large number of utterances, and thus only a portion of the symbols available are visible at any one time, with the communicator navigating the various pages. Speech-generating devices can produce electronic voice output by using digitized recordings of natural speech or through speech synthesis—which may carry less emotional information but can permit the user to speak novel messages. The content, organization, and updating of the vocabulary on an SGD is influenced by a number of factors, such as the user's needs and the contexts that the device will be used in. The development of techniques to improve the available vocabulary and rate of speech production is an active research area. Vocabulary items should be of high interest to the user, be frequently applicable, have a range of meanings, and be pragmatic in functionality. There are multiple methods of accessing messages on devices: directly or indirectly, or using specialized access devices—although the specific access method will depend on the skills and abilities of the user. SGD output is typically much slower than speech, although rate enhancement strategies can increase the user's rate of output, resulting in enhanced efficiency of communication. The first known SGD was prototyped in the mid-1970s, and rapid progress in hardware and software development has meant that SGD capabilities can now be integrated into devices like smartphones. Notable users of SGDs include Stephen Hawking, Roger Ebert, Tony Proudfoot, and Pete Frates (founder of the ALS Ice Bucket Challenge). Speech-generating systems may be dedicated devices developed solely for AAC, or non-dedicated devices such as computers running additional software to allow them to function as AAC devices. == History == SGDs have their roots in early electronic communication aids. The first such aid was a sip-and-puff typewriter controller named the patient-operated selector mechanism (Naman) prototyped by Reg Maling in the United Kingdom in 1960. POSSUM scanned through a set of symbols on an illuminated display. Researchers at Delft University in the Netherlands created the lightspot-operated typewriter (LOT) in 1970, which made use of small movements of the head to point a small spot of light at a matrix of characters, each equipped with a photoelectric cell. Although it was commercially unsuccessful, the LOT was well received by its users. In 1966, Barry Romich, a freshman engineering student at Case Western Reserve University, and Ed Prentke, an engineer at Highland View Hospital in Cleveland, Ohio, formed a partnership, creating the Prentke Romich Company. In 1969, the company produced its first communication device, a typing system based on a discarded Teletype machine. In 1979, Mark Dahmke developed software for a vocal communication aid program using the Computalker CT-1 analog speech synthesizer with a microcomputer. The software utilized phonemes to generate speech, assisting individuals with communication impairments in constructing words and sentences. Dahmke's work contributed to the advancement of assistive technology for people with disabilities. Notably, he designed the "Vocabulary Management System" for Bill Rush, a student with cerebral palsy. This early speech synthesis technology facilitated improved communication for Rush and was featured in a 1980 issue of LIFE Magazine. Dahmke's contributions have influenced the development of augmentative and alternative communication (AAC) technologies. During the 1970s and early 1980s, several other companies emerged that have since become prominent manufacturers of SGDs. Toby Churchill founded Toby Churchill Ltd in 1973, after losing his speech following encephalitis. In the US, Dynavox (then known as Sentient Systems Technology) grew out of a student project at Carnegie-Mellon University, created in 1982 to help a young woman with cerebral palsy to communicate. Beginning in the 1980s, improvements in technology led to a greatly increased number, variety, and performance of commercially available communication devices, and a reduction in their size and price. Alternative methods of access such as Target Scanning (also known as eye pointing) calibrate the movement of a user's eyes to direct an SGD to produce the desired speech. Scanning, in which alternatives are presented to the user sequentially, became available on communication devices. Speech output possibilities included both digitized and synthesized speech. Rapid progress in hardware and software development continued, including projects funded by the European Community. The first commercially available dynamic screen speech-generating devices were developed in the 1990s. Software was developed that allowed the computer-based production of communication boards. High-tech devices have continued to become smaller and lighter, while increasing accessibility and capability; communication devices can be accessed using eye-tracking systems, perform as a computer for word-processing and Internet use, and as an environmental control device for independent access to other equipment such as TV, radio and telephones. Stephen Hawking came to be associated with the unique voice of his particular synthesis equipment. Hawking was unable to speak due to a combination of disabilities caused by ALS, and an emergency tracheotomy. In the past 20 or so years SGD have gained popularity amongst young children with speech deficiencies, such as autism, Down syndrome, and predicted brain damage due to surgery. Starting in the early 2000s, specialists saw the benefit of using SGDs not only for adults but for children, as well. Neuro-linguists found that SGDs were just as effective in helping children who were at risk for temporary language deficits after undergoing brain surgery as it is for patients with ALS. In particular, digitized SGDs have been used as communication aids for pediatric patients during the recovery process. == Access methods == There are many methods of accessing messages on devices: directly, indirectly, and with specialized access devices. Direct access methods involve physical contact with the system, by using a keyboard or a touch screen. Users accessing SGDs indirectly and through specialized devices must manipulate an object in order to access the system, such as maneuvering a joystick, head mouse, optical head pointer, light pointer, infrared pointer, or switch access scanner. The specific access method will depend on the skills and abilities of the user. With direct selection a body part, pointer, adapted mouse, joystick, or eye tracking could be used, whereas switch access scanning is often used for indirect selection. Unlike direct selection (e.g., typing on a keyboard, touching a screen), users of Target Scanning can only make selections when the scanning indicator (or cursor) of the electronic device is on the desired choice. Those who are unable to point typically calibrate their eyes to use eye gaze as a way to point and blocking as a way to select desired words and phrases. The speed and pattern of scanning, as well as the way items are selected, are individualized to the physical, visual and cognitive capabilities of the user. == Message construction == Augmentative and alternative communication is typically much slower than speech, with users generally producing 8–10 words per minute. Rate enhancement strategies can increase the user's rate of output to around 12–15 words per minute, and as a result enhance the efficiency of communication. In any given SGD there may be a large number of vocal expressions that facilitate efficient and effective communication, including greetings, expressing desires, and asking questions. Some SGDs have multiple pages of symbols to accommodate a large number of vocal expressions, and thus only a portion of the symbols available are visible at any one time, with the communicator navigating the various pages. Speech-generating devices generally display a set of selections either using a dynamically changing screen, or a fixed display. There are two main options for increasing the rate of communication for an SGD: encoding, and prediction. Encoding permits a user to produce a word, sentence or phrase using only on
Video browsing
Video browsing, also known as exploratory video search, is the interactive process of skimming through video content in order to satisfy some information need or to interactively check if the video content is relevant. While originally proposed to help users inspecting a single video through visual thumbnails, modern video browsing tools enable users to quickly find desired information in a video archive by iterative human–computer interaction through an exploratory search approach. Many of these tools presume a smart user that wants features to interactively inspect video content, as well as automatic content filtering features. For that purpose, several video interaction features are usually provided, such as sophisticated navigation in video or search by a content-based query. Video browsing tools often build on lower-level video content analysis, such as shot transition detection, keyframe extraction, semantic concept detection, and create a structured content overview of the video file or video archive. Furthermore, they usually provide sophisticated navigation features, such as advanced timelines, visual seeker bars or a list of selected thumbnails, as well as means for content querying. Examples of content queries are shot filtering through visual concepts (e.g., only shots showing cars), through some specific characteristics (e.g., color or motion filtering), through user-provided sketches (e.g., a visually drawn sketch), or through content-based similarity search. == History == Video browsing was originally proposed by Iranian engineer Farshid Arman, Taiwanese computer scientist Arding Hsu, and computer scientist Ming-Yee Chiu, while working at Siemens, and it was presented at the ACM International Conference in August 1993. They described a shot detection algorithm for compressed video that was originally encoded with discrete cosine transform (DCT) video coding standards such as JPEG, MPEG and H.26x. The basic idea was that, since the DCT coefficients are mathematically related to the spatial domain and represent the content of each frame, they can be used to detect the differences between video frames. In the algorithm, a subset of blocks in a frame and a subset of DCT coefficients for each block are used as motion vector representation for the frame. By operating on compressed DCT representations, the algorithm significantly reduces the computational requirements for decompression and enables effective video browsing. The algorithm represents separate shots of a video sequence by an r-frame, a thumbnail of the shot framed by a motion tracking region. A variation of this concept was later adopted for QBIC video content mosaics, where each r-frame is a salient still from the shot it represents. === Video Notebook === Modern video browsing solutions include Video Notebook, a Menlo Park startup founded in 2021 by Mike Lanza, which uses computer vision to extract slides and optical character recognition and speech recognition to facilitate video search. The software can be either used on the client side (using a browser extension), where the slides and text are extracted while the video is watched (e.g. on a video platform like YouTube or Udemy), or on the server side. Processed videos, which can be viewed in the Video Notebook web app, feature a video browsing user interface with extracted timestamped slides, a search bar for querying the video (or a collection of videos), and text chapters. Video Notebook customers include organisations like Ernst & Young. === Video Browser Showdown === The Video Browser Showdown (VBS) is an annual live evaluation competition for exploratory video search tools, where international researchers use video browsing tools to solve ad-hoc video search tasks on a moderately large data set as fast as possible. The main goal of the VBS, which started in 2012 at the International Conference on MultiMedia Modeling (MMM), is to advance the performance of video browsing tools. Since 2016, the VBS also collaborates with TRECVID. The aim of the VBS is to evaluate video browsing tools for efficiency at known-item search (KIS) tasks with a well-defined data set in direct comparison to other tools.
Wayve
Wayve Technologies Ltd is a British autonomous driving technology company focused on developing self-driving vehicle systems through end-to-end deep learning. Founded in 2017 by researchers from the University of Cambridge, Wayve’s approach eschews detailed 3D maps and hand-coded rules, in favor of a self-learning “AI driver” that learns from camera data and driving experience. The London-headquartered startup has garnered significant attention and funding for its visually-based method. == History == Wayve was founded in Cambridge, England, on August 21, 2017, by Amar Shah and Alex Kendall, two machine learning PhD students at the University of Cambridge. Shah initially served as CEO while Kendall was CTO, and the pair set out to develop an unconventional self-driving car system using machine learning at every layer of the driving task. In May 2018, Wayve emerged from stealth mode with backing from early-stage investors. At this time the company had around 10 employees, and its advisory investors included Uber’s Chief Scientist, Zoubin Ghahramani, who shared Wayve’s vision of a learning-centric driving AI. In 2019, Wayve achieved a milestone by training a car to drive autonomously on public roads it had never seen before, using only cameras, a basic GPS map, and end-to-end deep learning control. The company moved its base to London and secured a $20 million Series A funding round in November 2019. This investment enabled Wayve to launch a pilot fleet of autonomous electric vehicles in central London for real-world testing. During these trials, Wayve’s cars (such as retrofitted Jaguar I-Pace SUVs) began navigating the complex, narrow streets of London to prove the system’s ability to adapt to challenging urban scenarios. In 2020, co-founder Amar Shah departed the company, and Alex Kendall assumed the role of CEO. The startup joined the Microsoft for Startups: Autonomous Driving program in 2020, leveraging Microsoft Azure’s cloud computing for training its machine learning models at scale. It also committed to testing exclusively on electric vehicles, and a goal to reduce carbon emissions. In 2021, Wayve entered pilot programs with major UK retailers. It launched a 12-month autonomous delivery trial with supermarket chain Asda, and received a £10 million ($13.6 million) investment from online grocer Ocado Group as part of a partnership to develop self-driving grocery delivery vans. Ocado’s backing gave Wayve access to a fleet of delivery vans for data collection and testing on busy London routes (with human safety drivers present) to train its AI in urban traffic. In 2022, after a successful Series B funding round, the company extended road testing beyond the UK to other regions, and, by 2023, in multiple countries. The company had begun operating in the United States and in continental Europe, in preparation for larger commercial deployments. In 2023, Wayve announced a collaboration with Nissan to integrate Wayve’s AI-driven software into its ProPilot ADAS system, slated to launch in fiscal year 2027. Wayve received strategic investment from Uber, in 2024, to jointly develop autonomous ride-hailing services. The two companies plan to trial a fully driverless robotaxi service in London, supported by a UK government program to accelerate commercial self-driving pilots to as early as 2026. To demonstrate the scalability of its technology, Wayve conducted an “AI-500” roadshow project, driving in dozens of cities across Asia, Europe, and North America using the same AI model. By mid-2025, it had completed autonomous driving demos in 90 cities without prior HD mapping. In April 2025, Wayve opened its first Asian research hub in Japan, with investment by SoftBank, to improve its model’s generalization using local driving data. That year, the company conducted driving tests in over 500 cities in Europe, North America and Japan without city-specific programming. In February 2026, Nissan, Uber and Wayve announced their collaboration on robotaxi development, with the aim of launching a pilot programme in Tokyo by late 2026. Wayve also formed a strategic alliance with Mercedes-Benz and Stellantis on personal vehicle and robotaxi applications. == Financing and investors == Wayve has been backed by a mix of venture capital (VC) firms, corporate investors, and individuals. Its initial seed funding came from funds such as Compound (NYC) and Firstminute Capital (London), as well as Cambridge-based angel investors, in 2018. Academic Pieter Abbeel and Uber’s chief scientist, Zoubin Ghahramani, were early backers. In November 2019, Wayve raised a $20 million Series A led by Eclipse Ventures, with participation from Balderton Capital and other prior investors. The Series A financing was used to fund the company’s first autonomous trials in London, and marked the first time a European self-driving car startup had secured a U.S. VC as lead investor. In October 2021, Ocado Group invested £10 million (approximately $13.6 million) in Wayve as a strategic partner in autonomous grocery delivery. This brought Wayve’s total funding to around $60 million at that time. The Series B round followed in January 2022, when Wayve announced $200 million in new funding led by Eclipse Ventures, with D1 Capital Partners, Moore Strategic Ventures, and Linse Capital. Balderton, Microsoft and Virgin Group joined as strategic backers. Baillie Gifford and Compound also participated; Ocado increased its stake as a strategic investor; and Meta AI head Yann LeCun and Richard Branson also became investors. Wayve’s Series C in May 2024 closed a $1.05 billion, led by Japan’s SoftBank Group. The funding round was the largest-ever for a UK AI company, and included new investor Nvidia, and returning investors Microsoft and Eclipse Ventures, among others. Uber also joined as a stratgic partner and a stakeholder. The Series C round increased Wayve’s total funding raised to about $1.3 billion to date from investors including SoftBank, Microsoft and Nvidia, and lifted Wayve’s valuation into “unicorn” status. In February 2026, Wayve announced a $1.2 billion Series D funding round; later that month, the company reported that $1.5 billion had been raised from, primarily, Mercedes-Benz, Stellantis, Nissan, and existing backers Uber, Microsoft and Nvidia, increasing Wayve's overall valuation to $8.6 billion. == Technology == Wayve’s self-driving approach centers on end-to-end deep learning and a vision-based AI system. Unlike conventional autonomous vehicles that depend on high-definition maps, hand-coded rules, and arrays of expensive lidar sensors, Wayve’s platform learns to drive predominantly using camera data and machine learning algorithms. The company refers to its AI-driven driving software as an “Embodied AI” or AI Driver, emphasizing that the system learns from experience (both real and simulated) to handle complex or novel situations rather than following pre-programmed instructions, not unlike Tesla's approach. The Wayve hardware-agnostic autonomy stack consists of a suite of video cameras, with basic automotive sensors, mounted on the vehicle, and paired with onboard compute units that are powered by GPUs to run the AI models. This vision-only philosophy is similar to Tesla’s Autopilot/FSDB model, but Wayve’s solution is vehicle-agnostic and mapless. Wayve’s strategy is to provide its driving AI as an OEM-ready platform; it plans to license or embed its technology into vehicles made by established automakers rather than build its own cars. Wayve’s development vehicles currently use Nvidia’s Orin system-on-chip as the onboard computer for running the AI model, but CEO Kendall has noted that the software can run on “whatever GPU [an automaker] already has in their vehicles” Wayve has built a cloud infrastructure, largely on Microsoft Azure, to process petabytes of this data, and uses simulation tools (known internally as the “Wayve Infinity” simulator) to synthetically generate and practice rare or dangerous scenarios for the AI to learn from. == Corporate affairs == Wayve is a privately held company headquartered in London, England, with its primary research and development office in the Kings Cross area of London. The company was initially incorporated as Wayve Technologies Ltd in the UK. Wayve has also established a presence in the U.S., in Silicon Valley); in Canada, with a research hub in Vancouver; in Yokohama, Japan; in Leonberg, Germany; and in Herzliya, Israel. The Leadership team includes research scientists and engineers with backgrounds in computer vision, robotics, and automotive systems. President Erez Dagan was hired in 2024, following two decades at Mobileye; chief scientist Jamie Shotton is formerly of Microsoft Research; CEO Alex Kendall, originally from New Zealand with a PhD in computer vision from Cambridge, took over as CEO in 2020 after the departure of his co-founder Amar Shah.