The artificial intelligence (AI) market in India is projected to reach $8 billion by 2025, growing at 40% CAGR from 2020 to 2025. This growth is part of the broader AI boom, a global period of rapid technological advancements with India being pioneer starting in the early 2010s with NLP based Chatbots from Haptik, Corover.ai, Niki.ai and then gaining prominence in the early 2020s based on reinforcement learning, marked by breakthroughs such as generative AI models from Krutrim, Sarvam, CoRover, OpenAI and Alphafold by Google DeepMind. In India, the development of AI has been similarly transformative, with applications in healthcare, finance, and education, bolstered by government initiatives like NITI Aayog's 2018 National Strategy for Artificial Intelligence. Institutions such as the Indian Statistical Institute and the Indian Institute of Science published breakthrough AI research papers and patents. India's transformation to AI is primarily being driven by startups and government initiatives & policies like Digital India. By fostering technological trust through digital public infrastructure, India is tackling socioeconomic issues by taking a bottom-up approach to AI. NASSCOM and Boston Consulting Group estimate that by 2027, India's AI services might be valued at $17 billion. According to 2025 Technology and Innovation Report, by UN Trade and Development, India ranks 10th globally for private sector investments in AI. According to Mary Meeker, India has emerged as a key market for AI platforms, accounting for the largest share of ChatGPT's mobile app users and having the third-largest user base for DeepSeek in 2025. While AI presents significant opportunities for economic growth and social development in India, challenges such as data privacy concerns, skill shortages, and ethical considerations need to be addressed for responsible AI deployment. The growth of AI in India has also led to an increase in the number of cyberattacks that use AI to target organizations. == History == === Early days (1960s-1980s) === The TIFRAC (Tata Institute of Fundamental Research Automatic Calculator) was designed and developed by a team led by Rangaswamy Narasimhan between 1954 and 1960. He worked on pattern recognition from 1961 to 1964 at the University of Illinois Urbana-Champaign's Digital Computer Laboratory. In order to conduct research on database technology, computer networking, computer graphics, and systems software, he and M. G. K. Menon founded the National Centre for Software Development and Computing Techniques. In 1965, he established the Computer Society of India and supervised the initial research work on AI at Tata Institute of Fundamental Research. Jagdish Lal launched the first computer science program in 1976 at Motilal Nehru Regional Engineering College. H. K. Kesavan from the University of Waterloo and Vaidyeswaran Rajaraman from the University of Wisconsin–Madison joined the IIT Kanpur Electrical Engineering Department in 1963–1964 as Assistant Professor and Head of Department, respectively. H.N. Mahabala, who was employed at Bendix Corporation's Computer Division, joined the department in 1965. He previously worked with Marvin Minsky. The IIT Kanpur Computer Center was led by H. K. Kesavan, with Vaidyeswaran Rajaraman serving as his deputy. Kesavan informally permitted Rajaraman and Mahabala to introduce artificial intelligence into computer science classes. The computer science program was approved by IIT Kanpur in 1971 and split out from the electrical engineering department. In 1973, an IBM System/370 Model 155 was installed at IIT Madras. John McCarthy, head of the Artificial Intelligence Laboratory at Stanford University visited IIT Kanpur in 1971. He donated PDP-1 with a time-sharing operating system. During the 1970s, the balance of payments deficit in India restricted import of computers. The Department of Computer Science and Automation at the Indian Institute of Science established in 1969, played an important role in nurturing the development of data science and artificial intelligence in India. First course on AI was introduced in the 1970s by G. Krishna. B. L. Deekshatulu introduced the first course on pattern recognition in the early 1970s. === Foundation phase === ==== 1980s ==== In the 1980s, the Indian Statistical Institute's Optical Character Recognition Project was one of the country's first attempts at studying artificial intelligence and machine learning. OCR technology has benefited greatly from the work of ISI's Computer Vision and Pattern Recognition Unit, which is headed by Bidyut Baran Chaudhuri. He also contributed in the development of computer vision and digital image processing. As part of the Indian Fifth Generation Computer Systems Research Programme, the Department of Electronics, with support from the United Nations Development Programme, initiated the Knowledge Based Computer Systems Project in 1986, marking the beginning of India's first major AI research program. Prime Minister Rajiv Gandhi requested that the Department of Electronics and IISc to initiate the Parallel Processing Project in 1986–1987. The Center for Development of Advanced Computing eventually joined those efforts. IIT Madras was selected to develop system diagnosis, ISI for image processing, National Centre for Software Technology for natural language processing and TIFR for speech processing. In 1987, the proposal of N. Seshagiri, Director General of the National Informatics Centre for the prototype development of supercomputer was cleared. Negotiations for a Cray supercomputer were underway between the Reagan administration and the Rajiv Gandhi government. US Defense Secretaries Frank Carlucci and Caspar Weinberger visited New Delhi after the US approved the transfer in 1988. The sale of a lower-end XMP-14 supercomputer was permitted in lieu of the Cray XMP-24 supercomputer due to security concerns. The Center for Development of Advanced Computing was formally established in March 1988 by the Ministry of Communications and Information Technology (previously the Ministry of IT) within the Department of Information Technology (formerly the Department of Electronics) in response to a recommendation made to the Prime Minister by the Scientific Advisory Council. The National Initiative in Supercomputing, which produced the PARAM series, was led by Vijay P. Bhatkar. For the first ten years, supercomputing and Indian language computing were the two main focus areas. C-DAC has expanded its operations in order to meet the needs in a number of domains, including network and internet software, real-time systems, artificial intelligence, and NLP. Under the direction of Professor KV Ramakrishnamacharyulu from National Sanskrit University and Professor Rajeev Sangal from the International Institute of Information Technology, Hyderabad, the Akshar Bharati Research Group was established in 1984 with support from IIT Kanpur and the University of Hyderabad for computational processing of Indian languages. They focused on computational linguistics, NLP with ontological database systems, and Indian language/translation theories with linguistic tradition. ==== 1990s ==== From IIT Kanpur, Mohan Tambe joined C-DAC in the 1990s to work on Graphics and Intelligence based Script Technology (GIST), which addressed the challenge of adapting personal computer software based on Latin script to Devanagiri and a number of other Indian language scripts. He was previously working on the Machine Translation for Indian languages Project. Within C-DAC, he established the GIST group. The technology was expanded to encompass NLP, artificial intelligence-based machine-aided language learning and translation, multimedia and multilingual computing solutions, and more. GIST resulted in the creation of G-CLASS (GIST cross language search plug-ins suite), a cross-language search engine. The Applied Artificial Intelligence Group at C-DAC has developed some basic and novel applications in the field of NLP, including machine translation, information extraction/retrieval, automatic summarization, speech recognition, text-to-speech synthesis, intelligent language teaching, and natural language-based document management with Decision Support Systems. These applications are the result of the foundation laid by previous language technology activities. Software firms in the Indian private sector began looking into AI applications, mostly in the area of business process automation. In order to allow machines to read, comprehend, and interpret human languages, the Language Technologies Research Center was founded in October 1999 at the International Institute of Information Technology, Hyderabad. It focused on the advancements in semantic parsing, information extraction, natural language generation, sentiment analysis, and dialogue systems. Some of the early AI research in India was driven by societal needs. For example; Eklavya, a knowledge-based program created by I
Personality computing
Personality computing is a research field related to artificial intelligence and personality psychology that studies personality by means of computational techniques from different sources, including text, multimedia, and social networks. == Overview == Personality computing addresses three main problems involving personality: automatic personality recognition, perception, and synthesis. Automatic personality recognition is the inference of the personality type of target individuals from their digital footprint. Automatic personality perception is the inference of the personality attributed by an observer to a target individual based on some observable behavior. Automatic personality synthesis is the generation of the style or behaviour of artificial personalities in Avatars and virtual agents. Self-assessed personality tests or observer ratings are always exploited as the ground truth for testing and validating the performance of artificial intelligence algorithms for the automatic prediction of personality types. There is a wide variety of personality tests, such as the Myers Briggs Type Indicator (MBTI) or the MMPI, but the most used are tests based on the Five Factor Model such as the Revised NEO Personality Inventory. Personality computing can be considered as an extension or complement of Affective computing, where the former focuses on personality traits and the latter on affective states. A further extension of the two fields is Character Computing which combines various character states and traits including but not limited to personality and affect. == History == Personality computing began around 2005 with the pioneering research in personality recognition by Shlomo Argamon and later by François Mairesse. These works showed that personality traits could be inferred with reasonable accuracy from text, such as blogs, self-presentations, and email addresses. In 2008, the concept of "portable personality" for the distributed management of personality profiles has been developed. A few years later, research began in personality recognition and perception from multimodal and social signals, such as recorded meetings and voice calls. In the 2010s, the research focused mainly on personality recognition and perception from social media, helped by the first workshops organized by Fabio Celli. In particular personality was extracted from Facebook, Twitter and Instagram. In the same years, automatic personality synthesis helped improve the coherence of simulated behavior in virtual agents. Scientific works by Michal Kosinski demonstrated the validity of Personality Computing from different digital footprints, in particular from user preferences such as Facebook page likes, showed that machines can recognize personality better than humans and raised a warning against Cambridge Analytica and misuse of this kind of technology. == Applications == Personality computing techniques, in particular personality recognition and perception, have applications in Social media marketing, where they can help reducing the cost of advertising campaigns through psychological targeting.
Exposure Notification
The (Google/Apple) Exposure Notification System (GAEN) is a framework and protocol specification developed by Apple Inc. and Google to facilitate digital contact tracing during the COVID-19 pandemic. When used by health authorities, it augments more traditional contact tracing techniques by automatically logging close approaches among notification system users using Android or iOS smartphones. Exposure Notification is a decentralized reporting protocol built on a combination of Bluetooth Low Energy technology and privacy-preserving cryptography. It is an opt-in feature within COVID-19 apps developed and published by authorized health authorities. Unveiled on April 10, 2020, it was made available on iOS on May 20, 2020, as part of the iOS 13.5 update and on December 14, 2020, as part of the iOS 12.5 update for older iPhones. On Android, it was added to devices via a Google Play Services update, supporting all versions since Android Marshmallow. The Apple/Google protocol is similar to the Decentralized Privacy-Preserving Proximity Tracing (DP-3T) protocol created by the European DP-3T consortium and the Temporary Contact Number (TCN) protocol by Covid Watch, but is implemented at the operating system level, which allows for more efficient operation as a background process. Since May 2020, a variant of the DP-3T protocol is supported by the Exposure Notification Interface. Other protocols are constrained in operation because they are not privileged over normal apps. This leads to issues, particularly on iOS devices where digital contact tracing apps running in the background experience significantly degraded performance. The joint approach is also designed to maintain interoperability between Android and iOS devices, which constitute nearly all of the market. The ACLU stated the approach "appears to mitigate the worst privacy and centralization risks, but there is still room for improvement". In late April, Google and Apple shifted the emphasis of the naming of the system, describing it as an "exposure notification service", rather than "contact tracing" system. == Technical specification == Digital contact tracing protocols typically have two major responsibilities: encounter logging and infection reporting. Exposure Notification only involves encounter logging which is a decentralized architecture. The majority of infection reporting is centralized in individual app implementations. To handle encounter logging, the system uses Bluetooth Low Energy to send tracking messages to nearby devices running the protocol to discover encounters with other people. The tracking messages contain unique identifiers that are encrypted with a secret daily key held by the sending device. These identifiers change every 15–20 minutes as well as Bluetooth MAC address in order to prevent tracking of clients by malicious third parties through observing static identifiers over time. The sender's daily encryption keys are generated using a random number generator. Devices record received messages, retaining them locally for 14 days. If a user tests positive for infection, the last 14 days of their daily encryption keys can be uploaded to a central server, where it is then broadcast to all devices on the network. The method through which daily encryption keys are transmitted to the central server and broadcast is defined by individual app developers. The Google-developed reference implementation calls for a health official to request a one-time verification code (VC) from a verification server, which the user enters into the encounter logging app. This causes the app to obtain a cryptographically signed certificate, which is used to authorize the submission of keys to the central reporting server. The received keys are then provided to the protocol, where each client individually searches for matches in their local encounter history. If a match meeting certain risk parameters is found, the app notifies the user of potential exposure to the infection. Google and Apple intend to use the received signal strength (RSSI) of the beacon messages as a source to infer proximity. RSSI and other signal metadata will also be encrypted to resist deanonymization attacks. === Version 1.0 === To generate encounter identifiers, first a persistent 32-byte private Tracing Key ( t k {\displaystyle tk} ) is generated by a client. From this a 16 byte Daily Tracing Key is derived using the algorithm d t k i = H K D F ( t k , N U L L , 'CT-DTK' | | D i , 16 ) {\displaystyle dtk_{i}=HKDF(tk,NULL,{\text{'CT-DTK'}}||D_{i},16)} , where H K D F ( Key, Salt, Data, OutputLength ) {\displaystyle HKDF({\text{Key, Salt, Data, OutputLength}})} is a HKDF function using SHA-256, and D i {\displaystyle D_{i}} is the day number for the 24-hour window the broadcast is in starting from Unix Epoch Time. These generated keys are later sent to the central reporting server should a user become infected. From the daily tracing key a 16-byte temporary Rolling Proximity Identifier is generated every 10 minutes with the algorithm R P I i , j = Truncate ( H M A C ( d t k i , 'CT-RPI' | | T I N j ) , 16 ) {\displaystyle RPI_{i,j}={\text{Truncate}}(HMAC(dtk_{i},{\text{'CT-RPI'}}||TIN_{j}),16)} , where H M A C ( Key, Data ) {\displaystyle HMAC({\text{Key, Data}})} is a HMAC function using SHA-256, and T I N j {\displaystyle TIN_{j}} is the time interval number, representing a unique index for every 10 minute period in a 24-hour day. The Truncate function returns the first 16 bytes of the HMAC value. When two clients come within proximity of each other they exchange and locally store the current R P I i , j {\displaystyle RPI_{i,j}} as the encounter identifier. Once a registered health authority has confirmed the infection of a user, the user's Daily Tracing Key for the past 14 days is uploaded to the central reporting server. Clients then download this report and individually recalculate every Rolling Proximity Identifier used in the report period, matching it against the user's local encounter log. If a matching entry is found, then contact has been established and the app presents a notification to the user warning them of potential infection. === Version 1.1 === Unlike version 1.0 of the protocol, version 1.1 does not use a persistent tracing key, rather every day a new random 16-byte Temporary Exposure Key ( t e k i {\displaystyle tek_{i}} ) is generated. This is analogous to the daily tracing key from version 1.0. Here i {\displaystyle i} denotes the time is discretized in 10 minute intervals starting from Unix Epoch Time. From this two 128-bit keys are calculated, the Rolling Proximity Identifier Key ( R P I K i {\displaystyle RPIK_{i}} ) and the Associated Encrypted Metadata Key ( A E M K i {\displaystyle AEMK_{i}} ). R P I K i {\displaystyle RPIK_{i}} is calculated with the algorithm R P I K i = H K D F ( t e k i , N U L L , 'EN-RPIK' , 16 ) {\displaystyle RPIK_{i}=HKDF(tek_{i},NULL,{\text{'EN-RPIK'}},16)} , and A E M K i {\displaystyle AEMK_{i}} using the algorithm A E M K i = H K D F ( t e k i , N U L L , 'EN-AEMK' , 16 ) {\displaystyle AEMK_{i}=HKDF(tek_{i},NULL,{\text{'EN-AEMK'}},16)} . From these values a temporary Rolling Proximity Identifier ( R P I i , j {\displaystyle RPI_{i,j}} ) is generated every time the BLE MAC address changes, roughly every 15–20 minutes. The following algorithm is used: R P I i , j = A E S 128 ( R P I K i , 'EN-RPI' | | 0 x 000000000000 | | E N I N j ) {\displaystyle RPI_{i,j}=AES128(RPIK_{i},{\text{'EN-RPI'}}||{\mathtt {0x000000000000}}||ENIN_{j})} , where A E S 128 ( Key, Data ) {\displaystyle AES128({\text{Key, Data}})} is an AES cryptography function with a 128-bit key, the data is one 16-byte block, j {\displaystyle j} denotes the Unix Epoch Time at the moment the roll occurs, and E N I N j {\displaystyle ENIN_{j}} is the corresponding 10-minute interval number. Next, additional Associated Encrypted Metadata is encrypted. What the metadata represents is not specified, likely to allow the later expansion of the protocol. The following algorithm is used: Associated Encrypted Metadata i , j = A E S 128 _ C T R ( A E M K i , R P I i , j , Metadata ) {\displaystyle {\text{Associated Encrypted Metadata}}_{i,j}=AES128\_CTR(AEMK_{i},RPI_{i,j},{\text{Metadata}})} , where A E S 128 _ C T R ( Key, IV, Data ) {\displaystyle AES128\_CTR({\text{Key, IV, Data}})} denotes AES encryption with a 128-bit key in CTR mode. The Rolling Proximity Identifier and the Associated Encrypted Metadata are then combined and broadcast using BLE. Clients exchange and log these payloads. Once a registered health authority has confirmed the infection of a user, the user's Temporary Exposure Keys t e k i {\displaystyle tek_{i}} and their respective interval numbers i {\displaystyle i} for the past 14 days are uploaded to the central reporting server. Clients then download this report and individually recalculate every Rolling Proximity Identifier starting from interval number i {\displaystyle i} ,
Automated restaurant
An automated restaurant or robotic restaurant is a restaurant that uses robots to do tasks such as delivering food and drink to the tables or cooking the food. Restaurant automation means the use of a restaurant management system to automate some or occasionally all of the major operations of a restaurant establishment. More recently, restaurants are opening that have completely or partially automated their services. These may include: taking orders, preparing food, serving, and billing. A few fully automated restaurants operate without any human intervention whatsoever. Robots are designed to help and sometimes replace human labour (such as waiters and chefs). The automation of restaurants may also allow for the option for greater customization of an order. == History == === Vending machines === In the late 19th and early 20th century a number of restaurants served food solely through vending machines. These restaurants were called automats or, in Japan, shokkenki. Customers ordered their food directly through the machines. === Sushi conveyors === Yoshiaki Shiraishi is a Japanese innovator who is known for the creation of conveyor belt sushi. He had the idea following difficulty staffing his small sushi restaurant and managing the restaurant on his own. He was inspired seeing beer bottles on a conveyor belt in an Asahi brewery. Yoshiaki's restaurants are an early example of restaurant automation; they used a conveyor belt to distribute dishes around the restaurant, eliminating the need for waiters. This example of automation dates back to the Japanese economic miracle; the first of Yoshiaki's conveyor belt sushi restaurants was opened under the name Mawaru Genroku Sushi in 1958, in Osaka. === Partial automation === As of 2011, across Europe, McDonald's had already begun implementing 7,000 touch screen kiosks that could handle cashiering duties. From 2015 to 2020, Zume had an automated pizza parlor. Later companies would try to produce smaller, less ambitious devices, with one robotics company producing a machine that could automate the slowest and most repetitive parts of assembling a pizza, such as spreading pizza sauce or placing slices of pepperoni, while leaving other customizations to employees. In 2020, a restaurant in the Netherlands began trialling the use of a robot to serve guests. In September 2021, Karakuri's 'Semblr' food service robot served personalised lunches for the 4,000 employees of grocery technology solutions provider ocado Group's head offices in Hatfield, UK. 2,700 different combinations of dishes were on offer. Customers could specify in grams what hot and cold items, proteins, sauces and fresh toppings they wanted. In 2021, Columbia University School of Engineering and Applied Science engineers developed a method of cooking 3D printed chicken with software-controlled robotic lasers. The “Digital Food” team exposed raw 3D printed chicken structures to both blue and infrared light. They then assessed the cooking depth, colour development, moisture retention and flavour differences of the laser-cooked 3D printed samples in comparison to stove-cooked meat. In June 2022 a California nonprofit chain of residential communities, Front Porch, experimented with robots in dining rooms at two locations to supplement wait staff by carrying plated food and drink to tables, and removing dishes. 65% of residents found the robots helpful, with 51% saying they let the staff spend more quality time with diners. 51% of staff were "excited" and 58% said they enabled more quality time with diners. The chain has 19 senior living communities (and 35 affordable housing communities), so it has potential to expand robots to more dining rooms. It is shifting to memory care, which may affect plans. == Rationales == === Advantages === Efficiency: Automated restaurants can significantly enhance operational efficiency by minimizing human error and reducing service time. With automated ordering, payment, and food preparation systems, customers can enjoy faster service and reduced waiting times. Cost savings: By reducing the need for human staff, automated restaurants can potentially lower labor costs. This can be particularly beneficial in areas with high labor expenses, as it allows for better resource allocation and cost management. Consistency: Automation ensures consistency in food quality and presentation. With precise portion control and standardized cooking methods, customers can expect the same quality and taste in their meals every time they visit. Enhanced customer experience: Self-service kiosks and automated systems provide customers with control and convenience. They can customize their orders, browse through menu options, and pay seamlessly, creating a more interactive and satisfying dining experience. === Disadvantages === Lack of personal touch: Automated restaurants may lack the personal interaction and warmth that traditional restaurants provide. Some customers prefer the human touch, personalized recommendations, and the social aspect of dining out. Technical issues: Reliance on technology means that technical glitches and malfunctions can occur, resulting in service disruptions or delays. Maintenance and technical support become critical in ensuring smooth operations. Limited menu complexity: The automation process may be better suited for standardized menu items rather than complex or customized dishes. The ability to cater to unique dietary preferences or accommodate special requests may be limited. Employment implications: Automated restaurants may result in job losses for traditional restaurant staff, potentially impacting the local workforce. It is important to consider the social and economic implications of adopting such technology. == Locations == Automated restaurants have been opening in many countries. Examples include: Nala Restaurant in Naperville, Illinois Fritz's Railroad Restaurant in Kansas City, Kansas Výtopna, a Railway Restaurant using model trains: franchise of various restaurants and coffeehouses in the Czech Republic Bagger's Restaurant in Nuremberg, Germany FuA-Men Restaurant, a ramen restaurant located in Nagoya, Japan Fōster Nutrition in Buenos Aires, Argentina Dalu Robot Restaurant in Jinan, China Haohai Robot Restaurant in Harbin, China Robot Kitchen Restaurant in Hong Kong Robo-Chef restaurant in Tehran, Iran, started in 2017, is the first robotic and "waiterless" restaurant of the Middle East. MIT graduates opened Spyce Kitchens in downtown Boston, Massachusetts, in 2018 Foodom, under Country Garden Holdings, opened January 12, 2020, in Guangzhou, China Robot Chacha, the first robot restaurant of India, is planning to open in the capital city of New Delhi. Kura Revolving Sushi Bar, with a number of locations in the United States, uses a tablets at tables for ordering, a conveyor belt to deliver food, and robots to deliver drinks and condiments. Chipotle Mexican Grill is beginning to deploy the Hyphen Makeline, which assembles up to 350 bowls and salads automatically per hour, and Chippy, an automatic tortilla chip fryer made by Miso Robotics. Serious Dumplings in Boca Raton, Florida
Oversampled binary image sensor
An oversampled binary image sensor is an image sensor with non-linear response capabilities reminiscent of traditional photographic film. Each pixel in the sensor has a binary response, giving only a one-bit quantized measurement of the local light intensity. The response function of the image sensor is non-linear and similar to a logarithmic function, which makes the sensor suitable for high dynamic range imaging. == Working principle == Before the advent of digital image sensors, photography, for the most part of its history, used film to record light information. At the heart of every photographic film are a large number of light-sensitive grains of silver-halide crystals. During exposure, each micron-sized grain has a binary fate: Either it is struck by some incident photons and becomes "exposed", or it is missed by the photon bombardment and remains "unexposed". In the subsequent film development process, exposed grains, due to their altered chemical properties, are converted to silver metal, contributing to opaque spots on the film; unexposed grains are washed away in a chemical bath, leaving behind the transparent regions on the film. Thus, in essence, photographic film is a binary imaging medium, using local densities of opaque silver grains to encode the original light intensity information. Thanks to the small size and large number of these grains, one hardly notices this quantized nature of film when viewing it at a distance, observing only a continuous gray tone. The oversampled binary image sensor is reminiscent of photographic film. Each pixel in the sensor has a binary response, giving only a one-bit quantized measurement of the local light intensity. At the start of the exposure period, all pixels are set to 0. A pixel is then set to 1 if the number of photons reaching it during the exposure is at least equal to a given threshold q. One way to build such binary sensors is to modify standard memory chip technology, where each memory bit cell is designed to be sensitive to visible light. With current CMOS technology, the level of integration of such systems can exceed 109~1010 (i.e., 1 giga to 10 giga) pixels per chip. In this case, the corresponding pixel sizes (around 50~nm ) are far below the diffraction limit of light, and thus the image sensor is oversampling the optical resolution of the light field. Intuitively, one can exploit this spatial redundancy to compensate for the information loss due to one-bit quantizations, as is classic in oversampling delta-sigma converters. Building a binary sensor that emulates the photographic film process was first envisioned by Fossum, who coined the name digital film sensor (now referred to as a quanta image sensor). The original motivation was mainly out of technical necessity. The miniaturization of camera systems calls for the continuous shrinking of pixel sizes. At a certain point, however, the limited full-well capacity (i.e., the maximum photon-electrons a pixel can hold) of small pixels becomes a bottleneck, yielding very low signal-to-noise ratios (SNRs) and poor dynamic ranges. In contrast, a binary sensor whose pixels need to detect only a few photon-electrons around a small threshold q has much less requirement for full-well capacities, allowing pixel sizes to shrink further. == Imaging model == === Lens === Consider a simplified camera model shown in Fig.1. The λ 0 ( x ) {\displaystyle \lambda _{0}(x)} is the incoming light intensity field. By assuming that light intensities remain constant within a short exposure period, the field can be modeled as only a function of the spatial variable x {\displaystyle x} . After passing through the optical system, the original light field λ 0 ( x ) {\displaystyle \lambda _{0}(x)} gets filtered by the lens, which acts like a linear system with a given impulse response. Due to imperfections (e.g., aberrations) in the lens, the impulse response, a.k.a. the point spread function (PSF) of the optical system, cannot be a Dirac delta, thus, imposing a limit on the resolution of the observable light field. However, a more fundamental physical limit is due to light diffraction. As a result, even if the lens is ideal, the PSF is still unavoidably a small blurry spot. In optics, such diffraction-limited spot is often called the Airy disk, whose radius R a {\displaystyle R_{a}} can be computed as R a = 1.22 w f , {\displaystyle R_{a}=1.22\,wf,} where w {\displaystyle w} is the wavelength of the light and f {\displaystyle f} is the F-number of the optical system. Due to the lowpass (smoothing) nature of the PSF, the resulting λ ( x ) {\displaystyle \lambda (x)} has a finite spatial-resolution, i.e., it has a finite number of degrees of freedom per unit space. === Sensor === Fig.2 illustrates the binary sensor model. The s m {\displaystyle s_{m}} denote the exposure values accumulated by the sensor pixels. Depending on the local values of s m {\displaystyle s_{m}} , each pixel (depicted as "buckets" in the figure) collects a different number of photons hitting on its surface. y m {\displaystyle y_{m}} is the number of photons impinging on the surface of the m {\displaystyle m} th pixel during an exposure period. The relation between s m {\displaystyle s_{m}} and the photon count y m {\displaystyle y_{m}} is stochastic. More specifically, y m {\displaystyle y_{m}} can be modeled as realizations of a Poisson random variable, whose intensity parameter is equal to s m {\displaystyle s_{m}} , As a photosensitive device, each pixel in the image sensor converts photons to electrical signals, whose amplitude is proportional to the number of photons impinging on that pixel. In a conventional sensor design, the analog electrical signals are then quantized by an A/D converter into 8 to 14 bits (usually the more bits the better). But in the binary sensor, the quantizer is 1 bit. In Fig.2, b m {\displaystyle b_{m}} is the quantized output of the m {\displaystyle m} th pixel. Since the photon counts y m {\displaystyle y_{m}} are drawn from random variables, so are the binary sensor output b m {\displaystyle b_{m}} . === Spatial and temporal oversampling === If it is allowed to have temporal oversampling, i.e., taking multiple consecutive and independent frames without changing the total exposure time τ {\displaystyle \tau } , the performance of the binary sensor is equivalent to the sensor with same number of spatial oversampling under certain condition. It means that people can make trade off between spatial oversampling and temporal oversampling. This is quite important, since technology usually gives limitation on the size of the pixels and the exposure time. == Advantages over traditional sensors == Due to the limited full-well capacity of conventional image pixel, the pixel will saturate when the light intensity is too strong. This is the reason that the dynamic range of the pixel is low. For the oversampled binary image sensor, the dynamic range is not defined for a single pixel, but a group of pixels, which makes the dynamic range high. == Reconstruction == One of the most important challenges with the use of an oversampled binary image sensor is the reconstruction of the light intensity λ ( x ) {\displaystyle \lambda (x)} from the binary measurement b m {\displaystyle b_{m}} . Maximum likelihood estimation can be used for solving this problem. Fig. 4 shows the results of reconstructing the light intensity from 4096 binary images taken by single photon avalanche diodes (SPADs) camera. A better reconstruction quality with fewer temporal measurements and faster, hardware friendly implementation, can be achieved by more sophisticated algorithms.
Convolutional neural network
A convolutional neural network (CNN) is a type of feedforward neural network that learns features via filter (or kernel) optimization. This type of deep learning network has been applied to process and make predictions from many different types of data including text, images and audio. CNNs are the de-facto standard in deep learning-based approaches to computer vision and image processing, and have only recently been replaced—in some cases—by newer architectures such as the transformer. Vanishing gradients and exploding gradients, seen during backpropagation in earlier neural networks, are prevented by the regularization that comes from using shared weights over fewer connections. For example, for each neuron in the fully-connected layer, 10,000 weights would be required for processing an image sized 100 × 100 pixels. However, applying cascaded convolution (or cross-correlation) kernels, only 25 weights for each convolutional layer are required to process 5x5-sized tiles. Higher-layer features are extracted from wider context windows, compared to lower-layer features. Some applications of CNNs include: image and video recognition, recommender systems, image classification, image segmentation, medical image analysis, natural language processing, brain–computer interfaces, and financial time series. CNNs are also known as shift invariant or space invariant artificial neural networks, based on the shared-weight architecture of the convolution kernels or filters that slide along input features and provide translation-equivariant responses known as feature maps. Counter-intuitively, most convolutional neural networks are not invariant to translation, due to the downsampling operation they apply to the input. Feedforward neural networks are usually fully connected networks, that is, each neuron in one layer is connected to all neurons in the next layer. The "full connectivity" of these networks makes them prone to overfitting data. Typical ways of regularization, or preventing overfitting, include: penalizing parameters during training (such as weight decay) or trimming connectivity (skipped connections, dropout, etc.) Robust datasets also increase the probability that CNNs will learn the generalized principles that characterize a given dataset rather than the biases of a poorly-populated set. Convolutional networks were inspired by biological processes in that the connectivity pattern between neurons resembles the organization of the animal visual cortex. Individual cortical neurons respond to stimuli only in a restricted region of the visual field known as the receptive field. The receptive fields of different neurons partially overlap such that they cover the entire visual field. CNNs use relatively little pre-processing compared to other image classification algorithms. This means that the network learns to optimize the filters (or kernels) through automated learning, whereas in traditional algorithms these filters are hand-engineered. This simplifies and automates the process, enhancing efficiency and scalability overcoming human-intervention bottlenecks. == Architecture == A convolutional neural network consists of an input layer, hidden layers and an output layer. In a convolutional neural network, the hidden layers include one or more layers that perform convolutions. Typically this includes a layer that performs a dot product of the convolution kernel with the layer's input matrix. This product is usually the Frobenius inner product, and its activation function is commonly ReLU. As the convolution kernel slides along the input matrix for the layer, the convolution operation generates a feature map, which in turn contributes to the input of the next layer. This is followed by other layers such as pooling layers, fully connected layers, and normalization layers. Here it should be noted how close a convolutional neural network is to a matched filter. === Convolutional layers === In a CNN, the input is a tensor with shape: (number of inputs) × (input height) × (input width) × (input channels) After passing through a convolutional layer, the image becomes abstracted to a feature map, also called an activation map, with shape: (number of inputs) × (feature map height) × (feature map width) × (feature map channels). Convolutional layers convolve the input and pass its result to the next layer. This is similar to the response of a neuron in the visual cortex to a specific stimulus. Each convolutional neuron processes data only for its receptive field. Although fully connected feedforward neural networks can be used to learn features and classify data, this architecture is generally impractical for larger inputs (e.g., high-resolution images), which would require massive numbers of neurons because each pixel is a relevant input feature. A fully connected layer for an image of size 100 × 100 has 10,000 weights for each neuron in the second layer. Convolution reduces the number of free parameters, allowing the network to be deeper. For example, using a 5 × 5 tiling region, each with the same shared weights, requires only 25 neurons. Using shared weights means there are many fewer parameters, which helps avoid the vanishing gradients and exploding gradients problems seen during backpropagation in earlier neural networks. To speed processing, standard convolutional layers can be replaced by depthwise separable convolutional layers, which are based on a depthwise convolution followed by a pointwise convolution. The depthwise convolution is a spatial convolution applied independently over each channel of the input tensor, while the pointwise convolution is a standard convolution restricted to the use of 1 × 1 {\displaystyle 1\times 1} kernels. === Pooling layers === Convolutional networks may include local and/or global pooling layers along with traditional convolutional layers. Pooling layers reduce the dimensions of data by combining the outputs of neuron clusters at one layer into a single neuron in the next layer. Local pooling combines small clusters, tiling sizes such as 2 × 2 are commonly used. Global pooling acts on all the neurons of the feature map. There are two common types of pooling in popular use: max and average. Max pooling uses the maximum value of each local cluster of neurons in the feature map, while average pooling takes the average value. === Fully connected layers === Fully connected layers connect every neuron in one layer to every neuron in another layer. It is the same as a traditional multilayer perceptron neural network (MLP). Each neuron in the fully connected layer receives input from all the neurons in the previous layer. These inputs are weighted and summed with the corresponding biases, and then passed through an activation function to perform a nonlinear transformation, generating the output. The flattened matrix goes through a fully connected layer to classify the images. === Receptive field === In neural networks, each neuron receives input from some number of locations in the previous layer. In a convolutional layer, each neuron receives input from only a restricted area of the previous layer called the neuron's receptive field. Typically the area is a square (e.g. 5 by 5 neurons). Whereas, in a fully connected layer, the receptive field is the entire previous layer. Thus, in each convolutional layer, each neuron takes input from a larger area in the input than previous layers. This is due to applying the convolution over and over, which takes the value of a pixel into account, as well as its surrounding pixels. When using dilated layers, the number of pixels in the receptive field remains constant, but the field is more sparsely populated as its dimensions grow when combining the effect of several layers. To manipulate the receptive field size as desired, there are some alternatives to the standard convolutional layer. For example, atrous or dilated convolution expands the receptive field size without increasing the number of parameters by interleaving visible and blind regions. Moreover, a single dilated convolutional layer can comprise filters with multiple dilation ratios, thus having a variable receptive field size. === Weights === Each neuron in a neural network computes an output value by applying a specific function to the input values received from the receptive field in the previous layer. The function that is applied to the input values is determined by a vector of weights and a bias (typically real numbers). Learning consists of iteratively adjusting these biases and weights. The vectors of weights and biases are called filters and represent particular features of the input (e.g., a particular shape). A distinguishing feature of CNNs is that many neurons can share the same filter. This reduces the memory footprint because a single bias and a single vector of weights are used across all receptive fields that share that filter, as opposed to each receptive field having its own bias and vector
Docic
Docic is a Tunisian digital health platform available as a web and mobile application, headquartered in Tunis, Tunisia. Founded in 2022 by Sami Kallel, an orthopedic surgeon, and Sofiane Trabelsi. The service helps patients and healthcare professionals store, organize, and share medical records digitally and to connect with the doctor online. == History == Docic was founded in 2022 as a health-technology company based in Tunisia, after which the mobile application was subsequently developed and made available to users. The platform was designed to provide healthcare professionals with access to patients’ complete medical history, including updates and recent changes, aiming at supporting clinical decision-making and reducing the risk of medical errors. In January 2025, Docic was listed amongst companies that have received the Startup Act label, which is a recognition under the Tunisian legal framework made to support innovative startups.