AI Image Generator From Image

AI Image Generator From Image — independent reviews, comparisons, pricing and step-by-step guides on Aizhi.

• 

  Image fusion

  The image fusion process is defined as gathering all the important information from multiple images, and their inclusion into fewer images, usually a single one. This single image is more informative and accurate than any single source image, and it consists of all the necessary information. The purpose of image fusion is not only to reduce the amount of data but also to construct images that are more appropriate and understandable for the human and machine perception. In computer vision, multisensor image fusion is the process of combining relevant information from two or more images into a single image. The resulting image will be more informative than any of the input images. In remote sensing applications, the increasing availability of space borne sensors gives a motivation for different image fusion algorithms. Several situations in image processing require high spatial and high spectral resolution in a single image. Most of the available equipment is not capable of providing such data convincingly. Image fusion techniques allow the integration of different information sources. The fused image can have complementary spatial and spectral resolution characteristics. However, the standard image fusion techniques can distort the spectral information of the multispectral data while merging. In satellite imaging, two types of images are available. The panchromatic image acquired by satellites is transmitted with the maximum resolution available and the multispectral data are transmitted with coarser resolution. This will usually be two or four times lower. At the receiver station, the panchromatic image is merged with the multispectral data to convey more information. Many methods exist to perform image fusion. The very basic one is the high-pass filtering technique. Later techniques are based on Discrete Wavelet Transform, uniform rational filter bank, and Laplacian pyramid. == Motivation == Multi sensor data fusion has become a discipline which demands more general formal solutions to a number of application cases. Several situations in image processing require both high spatial and high spectral information in a single image. This is important in remote sensing. However, the instruments are not capable of providing such information either by design or because of observational constraints. One possible solution for this is data fusion. == Methods == Image fusion methods can be broadly classified into two groups – spatial domain fusion and transform domain fusion. The fusion methods such as averaging, Brovey method, principal component analysis (PCA) and IHS based methods fall under spatial domain approaches. Another important spatial domain fusion method is the high-pass filtering based technique. Here the high frequency details are injected into upsampled version of MS images. The disadvantage of spatial domain approaches is that they produce spatial distortion in the fused image. Spectral distortion becomes a negative factor while we go for further processing, such as classification problem. Spatial distortion can be very well handled by frequency-domain approaches on image fusion. The multiresolution analysis has become a very useful tool for analysing remote sensing images. The discrete wavelet transform has become a very useful tool for fusion. Some other fusion methods are also there, such as Laplacian pyramid based, curvelet transform based etc. These methods show a better performance in spatial and spectral quality of the fused image compared to other spatial methods of fusion. The images used in image fusion should already be registered. Misregistration is a major source of error in image fusion. Some well-known image fusion methods are: High-pass filtering technique IHS transform based image fusion PCA-based image fusion Wavelet transform image fusion Pair-wise spatial frequency matching Comparative analysis of image fusion methods demonstrates that different metrics support different user needs, sensitive to different image fusion methods, and need to be tailored to the application. Categories of image fusion metrics are based on information theory features, structural similarity, or human perception. === Multi-focus image fusion === Multi-focus image fusion is used to collect useful and necessary information from input images with different focus depths in order to create an output image that ideally has all information from input images. In visual sensor network (VSN), sensors are cameras which record images and video sequences. In many applications of VSN, a camera can’t give a perfect illustration including all details of the scene. This is because of the limited depth of focus exists in the optical lens of cameras. Therefore, just the object located in the focal length of camera is focused and cleared and the other parts of image are blurred. VSN has an ability to capture images with different depth of focuses in the scene using several cameras. Due to the large amount of data generated by camera compared to other sensors such as pressure and temperature sensors and some limitation such as limited band width, energy consumption and processing time, it is essential to process the local input images to decrease the amount of transmission data. The aforementioned reasons emphasize the necessary of multi-focus images fusion. Multi-focus image fusion is a process which combines the input multi-focus images into a single image including all important information of the input images and it’s more accurate explanation of the scene than every single input image. == Applications == === In remote sensing === Image fusion in remote sensing has several application domains. An important domain is the multi-resolution image fusion (commonly referred to pan-sharpening). In satellite imagery we can have two types of images: Panchromatic images – An image collected in the broad visual wavelength range but rendered in black and white. Multispectral images – Images optically acquired in more than one spectral or wavelength interval. Each individual image is usually of the same physical area and scale but of a different spectral band. The SPOT PAN satellite provides high resolution (10m pixel) panchromatic data. While the LANDSAT TM satellite provides low resolution (30m pixel) multispectral images. Image fusion attempts to merge these images and produce a single high resolution multispectral image. The standard merging methods of image fusion are based on Red–Green–Blue (RGB) to Intensity–Hue–Saturation (IHS) transformation. The usual steps involved in satellite image fusion are as follows: Resize the low resolution multispectral images to the same size as the panchromatic image. Transform the R, G and B bands of the multispectral image into IHS components. Modify the panchromatic image with respect to the multispectral image. This is usually performed by histogram matching of the panchromatic image with Intensity component of the multispectral images as reference. Replace the intensity component by the panchromatic image and perform inverse transformation to obtain a high resolution multispectral image. Pan-sharpening can be done with Photoshop. Other applications of image fusion in remote sensing are available. === In medical imaging === Image fusion has become a common term used within medical diagnostics and treatment. The term is used when multiple images of a patient are registered and overlaid or merged to provide additional information. Fused images may be created from multiple images from the same imaging modality, or by combining information from multiple modalities, such as magnetic resonance image (MRI), computed tomography (CT), positron emission tomography (PET), and single-photon emission computed tomography (SPECT). In radiology and radiation oncology, these images serve different purposes. For example, CT images are used more often to ascertain differences in tissue density while MRI images are typically used to diagnose brain tumors. For accurate diagnosis, radiologists must integrate information from multiple image formats. Fused, anatomically consistent images are especially beneficial in diagnosing and treating cancer. With the advent of these new technologies, radiation oncologists can take full advantage of intensity modulated radiation therapy (IMRT). Being able to overlay diagnostic images into radiation planning images results in more accurate IMRT target tumor volumes.

  Read more →
• 

  InRule Technology

  InRule Technology is a software company that offers Business Rule Management System (BRMS) enterprise software products. == History == InRule Technology's Chief Executive Officer Rik Chomko and Chief Technology Officer Loren Goodman founded InRule Technology in Chicago in 2002. Paul Hessinger joined InRule Technology in 2004 as chief executive officer and chairman of the board and served until his retirement in 2015. They work with companies in several markets, including financial services, public sector, healthcare, and insurance. In 2007, InRule Technology became a charter member of the Microsoft Business Process Alliance. In August 2019, InRule was acquired by Open Gate Capital. == Products == On October 29, 2012, InRule Technology launched InRule for Microsoft Dynamics CRM. The program provides components to enable creation and update of rules within Microsoft Dynamics CRM, InRule for Microsoft Dynamics CRM provides a platform for shops that prefer to work with Microsoft's platforms. With the availability of InRule 4.6 in 2014, the company introduced deployment of InRule through REST services and allowed REST services to be called from InRule. This enables access to data exposed as a REST service and to package up a rule service for RESTful access. The product launch reflected the move of the company's core audience to use a broader array of technologies despite an earlier focus on .NET. In 2017, InRule introduced InRule for the Salesforce Platform, as well as a technology partnership with Work-Relay, a Business Process Management (BPM) application built on the Salesforce Platform. One year earlier the company introduced InRule for JavaScript, allowing enterprises to run rules on the client-side, server-side or both. The software architecture includes multiple components, including irAuthor, the primary authoring tool for creating and maintaining rules; irVerify, a real-time test environment to run and debug rule applications; and irSDK, a set of APIs that allows developers to integrate inRule into their applications. Additionally, irSOA allows users to access the InRule rule engine as a service. irSOA is now called the irServer Execution Service.

  Read more →
• 

  Modular Audio Recognition Framework

  Modular Audio Recognition Framework (MARF) is an open-source research platform and a collection of voice, sound, speech, text and natural language processing (NLP) algorithms written in Java and arranged into a modular and extensible framework that attempts to facilitate addition of new algorithms. MARF may act as a library in applications or be used as a source for learning and extension. A few example applications are provided to show how to use the framework. There is also a detailed manual and the API reference in the javadoc format as the project tends to be well documented. MARF, its applications, and the corresponding source code and documentation are released under the BSD-style license.

  Read more →
• 

  Lisp machine

  Lisp machines are general-purpose computers designed to efficiently run Lisp as their main software and programming language, usually via hardware support. They are an example of a high-level language computer architecture. In a sense, they were the first commercial single-user workstations. Despite being modest in number (perhaps 7,000 units total as of 1988) Lisp machines commercially pioneered some now-commonplace technologies, including networking innovations such as Chaosnet, and effective garbage collection. Several firms built and sold Lisp machines in the 1980s: Symbolics (3600, 3640, XL1200, MacIvory, and other models), Lisp Machines Incorporated (LMI Lambda), Texas Instruments (Explorer, MicroExplorer), and Xerox (Interlisp-D workstations). The operating systems were written in Lisp Machine Lisp, Interlisp (Xerox), and later partly in Common Lisp. == History == === Historical context === Artificial intelligence (AI) computer programs of the 1960s and 1970s intrinsically required what was then considered a huge amount of computer power, as measured in processor time and memory space. The power requirements of AI research were exacerbated by the Lisp symbolic programming language, when commercial hardware was designed and optimized for assembly- and Fortran-like programming languages. At first, the cost of such computer hardware meant that it had to be shared among many users. As integrated circuit technology shrank the size and cost of computers in the 1960s and early 1970s, and the memory needs of AI programs began to exceed the address space of the most common research computer, the Digital Equipment Corporation (DEC) PDP-10, researchers considered a new approach: a computer designed specifically to develop and run large artificial intelligence programs, and tailored to the semantics of the Lisp language. To provide consistent performance for interactive programs, these machines would often not be shared, but would be dedicated to a single user at a time. === Initial development === In 1973, Richard Greenblatt and Thomas Knight, programmers at Massachusetts Institute of Technology (MIT) Artificial Intelligence Laboratory (AI Lab), began what would become the MIT Lisp Machine Project when they first began building a computer hardwired to run certain basic Lisp operations, rather than run them in software, in a 24-bit tagged architecture. The machine also did incremental (or Arena) garbage collection. More specifically, since Lisp variables are typed at runtime rather than compile time, a simple addition of two variables could take five times as long on conventional hardware, due to test and branch instructions. Lisp Machines ran the tests in parallel with the more conventional single instruction additions. If the simultaneous tests failed, then the result was discarded and recomputed; this meant in many cases a speed increase by several factors. This simultaneous checking approach was used as well in testing the bounds of arrays when referenced, and other memory management necessities (not merely garbage collection or arrays). Type checking was further improved and automated when the conventional byte word of 32 bits was lengthened to 36 bits for Symbolics 3600-model Lisp machines and eventually to 40 bits or more (usually, the excess bits not accounted for by the following were used for error-correcting codes). The first group of extra bits were used to hold type data, making the machine a tagged architecture, and the remaining bits were used to implement compressed data representation (CDR) coding (wherein the usual linked list elements are compressed to occupy roughly half the space), aiding garbage collection by reportedly an order of magnitude. A further improvement was two microcode instructions which specifically supported Lisp functions, reducing the cost of calling a function to as little as 20 clock cycles, in some Symbolics implementations. The first machine was called the CONS machine (named after the list construction operator cons in Lisp). Often it was affectionately referred to as the Knight machine, perhaps since Knight wrote his master's thesis on the subject; it was extremely well received. It was subsequently improved into a version called CADR (a pun; in Lisp, the cadr function, which returns the second item of a list, is pronounced /ˈkeɪ.dəɹ/ or /ˈkɑ.dəɹ/, as some pronounce the word "cadre") which was based on essentially the same architecture. About 25 of what were essentially prototype CADRs were sold within and without MIT for ~$50,000; it quickly became the favorite machine for hacking – many of the most favored software tools were quickly ported to it (e.g. Emacs was ported from ITS in 1975). It was so well received at an AI conference held at MIT in 1978 that Defense Advanced Research Projects Agency (DARPA) began funding its development. === Commercializing MIT Lisp machine technology === In 1979, Russell Noftsker, being convinced that Lisp machines had a bright commercial future due to the strength of the Lisp language and the enabling factor of hardware acceleration, proposed to Greenblatt that they commercialize the technology. In a counter-intuitive move for an AI Lab hacker, Greenblatt acquiesced, hoping perhaps that he could recreate the informal and productive atmosphere of the Lab in a real business. These ideas and goals were considerably different from those of Noftsker. The two negotiated at length, but neither would compromise. As the proposed firm could succeed only with the full and undivided assistance of the AI Lab hackers as a group, Noftsker and Greenblatt decided that the fate of the enterprise was up to them, and so the choice should be left to the hackers. The ensuing discussions of the choice divided the lab into two factions. In February 1979, matters came to a head. The hackers sided with Noftsker, believing that a commercial venture-fund-backed firm had a better chance of surviving and commercializing Lisp machines than Greenblatt's proposed self-sustaining start-up. Greenblatt lost the battle. It was at this juncture that Symbolics, Noftsker's enterprise, slowly came together. While Noftsker was paying his staff a salary, he had no building or any equipment for the hackers to work on. He bargained with Patrick Winston that, in exchange for allowing Symbolics' staff to keep working out of MIT, Symbolics would let MIT use internally and freely all the software Symbolics developed. A consultant from CDC, who was trying to put together a natural language computer application with a group of West-coast programmers, came to Greenblatt, seeking a Lisp machine for his group to work with, about eight months after the disastrous conference with Noftsker. Greenblatt had decided to start his own rival Lisp machine firm, but he had done nothing. The consultant, Alexander Jacobson, decided that the only way Greenblatt was going to start the firm and build the Lisp machines that Jacobson desperately needed was if Jacobson pushed and otherwise helped Greenblatt launch the firm. Jacobson pulled together business plans, a board, a partner for Greenblatt (one F. Stephen Wyle). The newfound firm was named LISP Machine, Inc. (LMI), and was funded by CDC orders, via Jacobson. Around this time Symbolics (Noftsker's firm) began operating. It had been hindered by Noftsker's promise to give Greenblatt a year's head start, and by severe delays in procuring venture capital. Symbolics still had the major advantage that while 3 or 4 of the AI Lab hackers had gone to work for Greenblatt, 14 other hackers had signed onto Symbolics. Two AI Lab people were not hired by either: Richard Stallman and Marvin Minsky. Stallman, however, blamed Symbolics for the decline of the hacker community that had centered around the AI lab. For two years, from 1982 to the end of 1983, Stallman worked by himself to clone the output of the Symbolics programmers, with the aim of preventing them from gaining a monopoly on the lab's computers. Regardless, after a series of internal battles, Symbolics did get off the ground in 1980/1981, selling the CADR as the LM-2, while Lisp Machines, Inc. sold it as the LMI-CADR. Symbolics did not intend to produce many LM-2s, since the 3600 family of Lisp machines was supposed to ship quickly, but the 3600s were repeatedly delayed, and Symbolics ended up producing ~100 LM-2s, each of which sold for $70,000. Both firms developed second-generation products based on the CADR: the Symbolics 3600 and the LMI-LAMBDA (of which LMI managed to sell ~200). The 3600, which shipped a year late, expanded on the CADR by widening the machine word to 36-bits, expanding the address space to 28-bits, and adding hardware to accelerate certain common functions that were implemented in microcode on the CADR. The LMI-LAMBDA, which came out a year after the 3600, in 1983, was compatible with the CADR (it could run CADR microcode), but hardware differences existed. Texas Instruments (TI) joined the fray whe

  Read more →
• 

  Manifold regularization

  In machine learning, manifold regularization is a technique for using the shape of a dataset to constrain the functions that should be learned on that dataset. In many machine learning problems, the data to be learned do not cover the entire input space. For example, a facial recognition system may not need to classify any possible image, but only the subset of images that contain faces. The technique of manifold learning assumes that the relevant subset of data comes from a manifold, a mathematical structure with useful properties. The technique also assumes that the function to be learned is smooth: data with different labels are not likely to be close together, and so the labeling function should not change quickly in areas where there are likely to be many data points. Because of this assumption, a manifold regularization algorithm can use unlabeled data to inform where the learned function is allowed to change quickly and where it is not, using an extension of the technique of Tikhonov regularization. Manifold regularization algorithms can extend supervised learning algorithms in semi-supervised learning and transductive learning settings, where unlabeled data are available. The technique has been used for applications including medical imaging, geographical imaging, and object recognition. == Manifold regularizer == === Motivation === Manifold regularization is a type of regularization, a family of techniques that reduces overfitting and ensures that a problem is well-posed by penalizing complex solutions. In particular, manifold regularization extends the technique of Tikhonov regularization as applied to Reproducing kernel Hilbert spaces (RKHSs). Under standard Tikhonov regularization on RKHSs, a learning algorithm attempts to learn a function f {\displaystyle f} from among a hypothesis space of functions H {\displaystyle {\mathcal {H}}} . The hypothesis space is an RKHS, meaning that it is associated with a kernel K {\displaystyle K} , and so every candidate function f {\displaystyle f} has a norm ‖ f ‖ K {\displaystyle \left\|f\right\|_{K}} , which represents the complexity of the candidate function in the hypothesis space. When the algorithm considers a candidate function, it takes its norm into account in order to penalize complex functions. Formally, given a set of labeled training data ( x 1 , y 1 ) , … , ( x ℓ , y ℓ ) {\displaystyle (x_{1},y_{1}),\ldots ,(x_{\ell },y_{\ell })} with x i ∈ X , y i ∈ Y {\displaystyle x_{i}\in X,y_{i}\in Y} and a loss function V {\displaystyle V} , a learning algorithm using Tikhonov regularization will attempt to solve the expression arg min f ∈ H 1 ℓ ∑ i = 1 ℓ V ( f ( x i ) , y i ) + γ ‖ f ‖ K 2 {\displaystyle {\underset {f\in {\mathcal {H}}}{\arg \!\min }}{\frac {1}{\ell }}\sum _{i=1}^{\ell }V(f(x_{i}),y_{i})+\gamma \left\|f\right\|_{K}^{2}} where γ {\displaystyle \gamma } is a hyperparameter that controls how much the algorithm will prefer simpler functions over functions that fit the data better. Manifold regularization adds a second regularization term, the intrinsic regularizer, to the ambient regularizer used in standard Tikhonov regularization. Under the manifold assumption in machine learning, the data in question do not come from the entire input space X {\displaystyle X} , but instead from a nonlinear manifold M ⊂ X {\displaystyle M\subset X} . The geometry of this manifold, the intrinsic space, is used to determine the regularization norm. === Laplacian norm === There are many possible choices for the intrinsic regularizer ‖ f ‖ I {\displaystyle \left\|f\right\|_{I}} . Many natural choices involve the gradient on the manifold ∇ M {\displaystyle \nabla _{M}} , which can provide a measure of how smooth a target function is. A smooth function should change slowly where the input data are dense; that is, the gradient ∇ M f ( x ) {\displaystyle \nabla _{M}f(x)} should be small where the marginal probability density P X ( x ) {\displaystyle {\mathcal {P}}_{X}(x)} , the probability density of a randomly drawn data point appearing at x {\displaystyle x} , is large. This gives one appropriate choice for the intrinsic regularizer: ‖ f ‖ I 2 = ∫ x ∈ M ‖ ∇ M f ( x ) ‖ 2 d P X ( x ) {\displaystyle \left\|f\right\|_{I}^{2}=\int _{x\in M}\left\|\nabla _{M}f(x)\right\|^{2}\,d{\mathcal {P}}_{X}(x)} In practice, this norm cannot be computed directly because the marginal distribution P X {\displaystyle {\mathcal {P}}_{X}} is unknown, but it can be estimated from the provided data. === Graph-based approach of the Laplacian norm === When the distances between input points are interpreted as a graph, then the Laplacian matrix of the graph can help to estimate the marginal distribution. Suppose that the input data include ℓ {\displaystyle \ell } labeled examples (pairs of an input x {\displaystyle x} and a label y {\displaystyle y} ) and u {\displaystyle u} unlabeled examples (inputs without associated labels). Define W {\displaystyle W} to be a matrix of edge weights for a graph, where W i j {\displaystyle W_{ij}} is a similarity built from distance measure between the data points x i {\displaystyle x_{i}} and x j {\displaystyle x_{j}} (so that more close implies higher W i j {\displaystyle W_{ij}} ). Define D {\displaystyle D} to be a diagonal matrix with D i i = ∑ j = 1 ℓ + u W i j {\displaystyle D_{ii}=\sum _{j=1}^{\ell +u}W_{ij}} and L {\displaystyle L} to be the Laplacian matrix D − W {\displaystyle D-W} . Then, as the number of data points ℓ + u {\displaystyle \ell +u} increases, L {\displaystyle L} converges to the Laplace–Beltrami operator Δ M {\displaystyle \Delta _{M}} , which is the divergence of the gradient ∇ M {\displaystyle \nabla _{M}} . Then, if f {\displaystyle \mathbf {f} } is a vector of the values of f {\displaystyle f} at the data, f = [ f ( x 1 ) , … , f ( x l + u ) ] T {\displaystyle \mathbf {f} =[f(x_{1}),\ldots ,f(x_{l+u})]^{\mathrm {T} }} , the intrinsic norm can be estimated: ‖ f ‖ I 2 = 1 ( ℓ + u ) 2 f T L f {\displaystyle \left\|f\right\|_{I}^{2}={\frac {1}{(\ell +u)^{2}}}\mathbf {f} ^{\mathrm {T} }L\mathbf {f} } As the number of data points ℓ + u {\displaystyle \ell +u} increases, this empirical definition of ‖ f ‖ I 2 {\displaystyle \left\|f\right\|_{I}^{2}} converges to the definition when P X {\displaystyle {\mathcal {P}}_{X}} is known. === Solving the regularization problem with graph-based approach === Using the weights γ A {\displaystyle \gamma _{A}} and γ I {\displaystyle \gamma _{I}} for the ambient and intrinsic regularizers, the final expression to be solved becomes: arg min f ∈ H 1 ℓ ∑ i = 1 ℓ V ( f ( x i ) , y i ) + γ A ‖ f ‖ K 2 + γ I ( ℓ + u ) 2 f T L f {\displaystyle {\underset {f\in {\mathcal {H}}}{\arg \!\min }}{\frac {1}{\ell }}\sum _{i=1}^{\ell }V(f(x_{i}),y_{i})+\gamma _{A}\left\|f\right\|_{K}^{2}+{\frac {\gamma _{I}}{(\ell +u)^{2}}}\mathbf {f} ^{\mathrm {T} }L\mathbf {f} } As with other kernel methods, H {\displaystyle {\mathcal {H}}} may be an infinite-dimensional space, so if the regularization expression cannot be solved explicitly, it is impossible to search the entire space for a solution. Instead, a representer theorem shows that under certain conditions on the choice of the norm ‖ f ‖ I {\displaystyle \left\|f\right\|_{I}} , the optimal solution f ∗ {\displaystyle f^{}} must be a linear combination of the kernel centered at each of the input points: for some weights α i {\displaystyle \alpha _{i}} , f ∗ ( x ) = ∑ i = 1 ℓ + u α i K ( x i , x ) {\displaystyle f^{}(x)=\sum _{i=1}^{\ell +u}\alpha _{i}K(x_{i},x)} Using this result, it is possible to search for the optimal solution f ∗ {\displaystyle f^{}} by searching the finite-dimensional space defined by the possible choices of α i {\displaystyle \alpha _{i}} . === Functional approach of the Laplacian norm === The idea beyond the graph-Laplacian is to use neighbors to estimate the Laplacian. This method is akin to local averaging methods, that are known to scale poorly in high-dimensional problems. Indeed, the graph Laplacian is known to suffer from the curse of dimensionality. Luckily, it is possible to leverage expected smoothness of the function to estimate thanks to more advanced functional analysis. This method consists of estimating the Laplacian operator using derivatives of the kernel reading ∂ 1 , j K ( x i , x ) {\displaystyle \partial _{1,j}K(x_{i},x)} where ∂ 1 , j {\displaystyle \partial _{1,j}} denotes the partial derivatives according to the j-th coordinate of the first variable. This second approach to the Laplacian norm is to put in relation with meshfree methods, that contrast with the finite difference method in PDE. == Applications == Manifold regularization can extend a variety of algorithms that can be expressed using Tikhonov regularization, by choosing an appropriate loss function V {\displaystyle V} and hypothesis space H {\displaystyle {\mathcal {H}}} . Two commonly used examples are the families of support vector machines and regularized least squares algorithm

  Read more →
• 

  Quantum Artificial Intelligence Lab

  The Quantum Artificial Intelligence Lab (also called the Quantum AI Lab or QuAIL) is a joint initiative of NASA, Universities Space Research Association, and Google (specifically, Google Research) whose goal is to pioneer research on how quantum computing might help with machine learning and other difficult computer science problems. The lab is hosted at NASA's Ames Research Center. == History == The Quantum AI Lab was announced by Google Research in a blog post on May 16, 2013. At the time of launch, the Lab was using the most advanced commercially available quantum computer, D-Wave Two from D-Wave Systems. On October 10, 2013, Google released a short film describing the current state of the Quantum AI Lab. On October 18, 2013, Google announced that it had incorporated quantum physics into Minecraft. In January 2014, Google reported results comparing the performance of the D-Wave Two in the lab with that of classical computers. The results were ambiguous and provoked heated discussion on the Internet. On 2 September 2014, it was announced that the Google Quantum AI Lab, in partnership with UC Santa Barbara, would be launching an initiative to create quantum information processors based on superconducting electronics. On the 23rd of October 2019, the Quantum AI Lab announced in a paper that it had achieved quantum supremacy with their Sycamore processor. The claim of quantum supremacy achievement has since been debated, with a far more accurate simulation on a classical computer being possible in 2.5 days as a conservative estimate. == Present == On December 9, 2024, Google introduced the Willow processor, describing it as a "state-of-the-art quantum chip". Google claims that this new chip takes just five minutes to solve a problem that takes traditional supercomputers ten septillion years. However, experts say Willow is, for now, a largely experimental device.

  Read more →
• 

  Turing's Wager

  Turing's Wager is a philosophical argument that claims it is impossible to infer or deduce a detailed mathematical model of the human brain within a reasonable timescale, and thus impossible in any practical sense. The argument was first given in 1950 by the computational theorist Alan Turing in his paper Computing Machinery and Intelligence, published in Mind (Turing 1950, p. 453). The argument asserts that determining any mathematical model of a computer (its source code or any isomorphic equivalent such as a Turing machine or virtual simulation) is not possible in a reasonable timeframe. As a consequence, determining a mathematical model of the human brain (which is, by its nature, more complicated) must also be impossible within that timeframe. == Effect of modern technology on the wager == It has been argued that modern neuroimaging techniques will allow researchers to create accurate simulations of the human mind within the 21st century (Kurzweil 2012; Markram 2012, Fildes 2009), thereby overcoming the wager. Others have argued that such claims are unjustified (Thwaites et al. 2017). == Relationship between Turing's Wager and the Turing Test == The Turing Test attempts to define when a machine might be said to possess human intelligence, while Turing's Wager is an argument aiming to demonstrate that characterising the brain mathematically will take over a thousand years. While building an artificial intelligence and mapping the human brain are both difficult endeavours, the former is actually a sub-problem of the latter (Thwaites et al. 2017).

  Read more →
• 

  The Emperor's New Mind

  The Emperor's New Mind: Concerning Computers, Minds and The Laws of Physics is a 1989 book by the mathematical physicist Roger Penrose that posits a quantum mind theory. Penrose argues that human consciousness is non-algorithmic, and thus is not capable of being modeled by a conventional Turing machine, which includes a digital computer. Penrose hypothesizes that quantum mechanics plays an essential role in the understanding of human consciousness. The collapse of the quantum wavefunction is seen as playing an important role in brain function. Most of the book is spent reviewing, for the scientifically-minded lay-reader, a plethora of interrelated subjects such as Newtonian physics, special and general relativity, the philosophy and limitations of mathematics, quantum physics, cosmology, and the nature of time. Penrose intermittently describes how each of these bears on his developing theme: that consciousness is not "algorithmic". Only the later portions of the book address the thesis directly. == Overview == Penrose states that his ideas on the nature of consciousness are speculative, and his thesis is considered erroneous by some experts in the fields of philosophy, computer science, and robotics. The Emperor's New Mind attacks the claims of artificial intelligence using the physics of computing: Penrose notes that the present home of computing lies more in the tangible world of classical mechanics than in the imponderable realm of quantum mechanics. The modern computer is a deterministic system that for the most part simply executes algorithms. Penrose shows that, by reconfiguring the boundaries of a billiard table, one might make a computer in which the billiard balls act as message carriers and their interactions act as logical decisions. The billiard-ball computer was first designed some years ago by Edward Fredkin and Tommaso Toffoli of the Massachusetts Institute of Technology. == Reception == Following the publication of the book, Penrose began to collaborate with Stuart Hameroff on a biological analog to quantum computation involving microtubules, which became the foundation for his subsequent book, Shadows of the Mind: A Search for the Missing Science of Consciousness. Penrose won the Science Book Prize in 1990 for The Emperor's New Mind. According to an article in the American Journal of Physics, Penrose incorrectly claims a barrier far away from a localized particle can affect the particle.

  Read more →
• 

  I Am Rich

  I Am Rich is a discontinued 2008 mobile app for iPhones which had minimal function and was priced at US$999.99 (equivalent to $1,495 in 2025). The app was pulled from the App Store less than 24 hours after its launch. Receiving negative reviews from critics, only eight copies were sold. In the years since, several similar applications have been released at lower prices. == Overview == I Am Rich was developed as a joke by German software developer, Armin Heinrich, after he saw iPhone users complaining about software priced above $0.99. The app only showed a glowing red gem and an icon that, when pressed, displayed the following mantra in large text: I am richI deserv [sic] itI am good,healthy & successful Heinrich told The New York Times that "I regard it as art. I did not expect many people to buy it and did not expect all the fuss about it." The application is described as "a work of art with no hidden function at all", with its only purpose being to show other people that they were able to afford it. Vox writer Zachary Crockett called it "the ultimate Veblen good in app form". == Release == Heinrich released and distributed I Am Rich through the App Store on 5 August 2008. The app was sold for US$999.99 (equivalent to $1,495 in 2025), €799.99 (equivalent to €1,078 in 2023), and £599.99 (equivalent to £978.12 in 2025)—the highest prices Apple allowed for App Store content. Without explanation, the application was removed from the App Store by Apple less than a day after its release. === Purchases === Eight people bought the application, at least one of whom claimed to have done so accidentally. Six US sales and two European sales netted $5,600 for Heinrich and $2,400 for Apple (respectively equivalent to $8,374 and $3,589 in 2025). In correspondence with the Los Angeles Times, Heinrich told the newspaper that Apple had refunded two purchasers of his app, and that he was happy to not have dissatisfied customers. == Reception == Discussing the app on the website Silicon Alley Insider, Dan Frommer described the program as a "scam", "worthless", and finally "a joke that smells like a scammy rip-off" on August 5, 6, and 8, respectively. Without purchasing the app, Fox News's Paul Wagenseil guessed that the secret mantra was "German for 'Sucker!'" (Heinrich is German). Wired's Brian X. Chen described I Am Rich as a waste of money to "prove you're a jerk", and contrasted the expenditure with donating to cancer foundations and Third World countries. Heinrich told the Los Angeles Times's Mark Milian that he had received correspondence from satisfied customers: "I've got e-mails from customers telling me that they really love the app [... and that they had] no trouble spending the money". In an interview with The New York Times, though, he told of receiving many insulting emails and telephone messages. == Similar applications == The next year, Heinrich released I Am Rich LE. Priced at US$9.99 (equivalent to $14.99 in 2025), the new app has several new features (including a calculator, "help system", and the "famous mantra without the spelling mistakes") to meet Apple's requirement that apps have "definable content". Some customers were disappointed by the new functionality, poorly rating the app due to its ostensible improvements. On 23 February 2009, CNET Asia reported on the "conceptually similar" app, I Am Richer, developed by Mike DG for Google's Android. The app was released on the Android Market for US$200 (equivalent to $300.14 in 2025), a limit imposed by Google, who had no objection to the application. With the same name, the I Am Rich that was released on the Windows Phone Marketplace on 22 December 2010, was developed by DotNetNuzzi. Described by MobileCrunch as equally useless as the original, this app cost US$499.99 (equivalent to $738.2 in 2025), the price cap imposed by Microsoft.

  Read more →
• 

  IEEE Transactions on Pattern Analysis and Machine Intelligence

  IEEE Transactions on Pattern Analysis and Machine Intelligence (sometimes abbreviated as IEEE PAMI or simply PAMI) is a monthly peer-reviewed scientific journal published by the IEEE Computer Society. == Background == The journal covers research in computer vision and image understanding, pattern analysis and recognition, machine intelligence, machine learning, search techniques, document and handwriting analysis, medical image analysis, video and image sequence analysis, content-based retrieval of image and video, and face and gesture recognition. The editor-in-chief is Kyoung Mu Lee (Seoul National University). According to the Journal Citation Reports, the journal has a 2023 impact factor of 20.8.

  Read more →
• 

  Google Nest

  Google Nest, formerly branded Google Home, is a line of smart home products including smart speakers, smart displays, streaming devices, thermostats, smoke detectors, routers and security systems including smart doorbells, cameras and smart locks. The Nest brand name was originally owned by Nest Labs, co-founded by former Apple engineers Tony Fadell and Matt Rogers in 2010. Its flagship product, which was the company's first offering, is the Nest Learning Thermostat, introduced in 2011. The product is programmable, self-learning, sensor-driven, and Wi-Fi-enabled: features that are often found in other Nest products. It was followed by the Nest Protect smoke and carbon monoxide detectors in October 2013. After its acquisition of Dropcam in 2014, the company introduced its Nest Cam branding of security cameras beginning in June 2015. The company quickly expanded to more than 130 employees by the end of 2012. Google acquired Nest Labs for US$3.2 billion in January 2014, when the company employed 280. As of late 2015, Nest employs more than 1,100 and added a primary engineering center in Seattle. After Google reorganized itself under the holding company Alphabet Inc., Nest operated independently of Google from 2015 to 2018. However, in 2018, Nest was merged into Google's home-devices unit led by Rishi Chandra, effectively ceasing to exist as a separate business. In July 2018, it was announced that all Google Home electronics products will henceforth be marketed under the brand Google Nest. == History == === Nest Labs before acquisition by Google === Nest Labs was founded in 2010 by former Apple engineers Tony Fadell and Matt Rogers. The idea came when Fadell was building a vacation home and found all of the available thermostats on the market to be inadequate, motivated to bring something better on the market. Early investors in Nest Labs included Shasta Ventures and Kleiner Perkins. === Acquisition by Google of Nest Labs, Dropcam, and Revolv === On January 13, 2014, Google announced plans to acquire Nest Labs for $3.2 billion in cash. Google completed the acquisition the next day, on January 14, 2014. The company would operate independently from Google's other businesses. In June 2014, it was announced that Nest would buy camera startup Dropcam for $555 million. With the purchase, Dropcam became integrated with other Nest products; if the Protect alarm is triggered, the Dropcam can automatically start recording, and the Thermostat can use Dropcam to sense for motion. In September 2014, the Nest Thermostat and Nest Protect (a smoke alarm) became available in Belgium, France, Ireland, and the Netherlands. Initially, they were sold in approximately 400 stores across Europe, with another 150 stores to be added by the end of the year. In June 2015, the new Nest Cam, replacing the Dropcam, was announced, together with the second generation of the Nest Protect; there were internal reports that sales of the rebranded camera fell. On October 24, 2014, Nest both acquired the hub service Revolv, and discontinued its product line, gaining the expertise of Revolv's staff. === Nest as a subsidiary of Alphabet Inc. === In August 2015, Google announced that it would restructure its operations under a new parent company, Alphabet Inc., with Nest being separated from Google as a subsidiary of the new holding company. In January 2016, some Nest thermostats stopped working, a fault attributed to a software update from two weeks earlier. There were no lawsuits, individual or class-action, due to an arbitration clause in the contract. All Revolv smart hubs, costing several hundred dollars, were deliberately remotely bricked on May 15, 2016; notice was posted on the company's website in February. The story became news on April 4. The "lifetime subscription" to Revolv's online service, which had been sold with the hub, was defined by Nest to be the lifetime of the device, which ended May 15. Nest's decision to brick the hubs, and its "acerbic" corporate culture, faced substantial criticism from within Google/Alphabet and in press coverage. Many of Nest's staffers came from Dropcam and Revolv, and by November 2015, about 70 of about 1000 staffers had quit, causing management concern. Some countermeasures had been taken in takeover deals, to financially discourage senior people from leaving before set dates. Of the ~100 Dropcam staffers, about half had left by March 2016, when former Dropcam CEO Greg Duffy (who left 8 months after the takeover) wrote a post openly regretting selling his company to Nest. He stated that about 500 people had left (of a 1200-person staff). On June 6, 2016, Tony Fadell, the Nest CEO, announced in a blog post that he was leaving the company he founded with Matt Rogers and stepping into an "advisory" role. At this point the Nest acquisition was described by some press as a "disaster" for Google. As of mid-June 2016, Nest's problems were considered symptomatic of the limited market for home automation. According to Frank Gillet of Forrester Research, only 6% of American households possessed internet-connected devices such as appliances, home-monitoring systems, speakers, or lighting. He also predicted this percentage would grow to only 15% by 2021. Furthermore, 72% of respondents in a 2016 British survey conducted by Pricewaterhouse Coopers did not foresee adopting smart-home technology over the next two to five years. === Nest as a part of Google hardware division === On February 7, 2018, it was announced by hardware head Rick Osterloh that Nest had been merged into Google's hardware division, directly alongside units such as Google Home and Chromecast. It would retain its separate Palo Alto headquarters, but Nest CEO Marwan Fawaz would now report to Osterloh, and there were plans for tighter integration with Google platforms and software such as Google Assistant in future products. Shortly after the announcement, co-founder and chief product officer Matt Rogers announced his plans to leave the company. On July 18, 2018, Nest CEO Marwan Fawaz stepped down. Nest was merged with Google's home devices team, led by Rishi Chandra. During the Google I/O keynote on May 7, 2019, it was announced that Google Nest will now serve as the blanket branding for all of Google's home products. The Google Home Hub was retroactively renamed Google Nest Hub, while a new and larger version of the product is now available called the Nest Hub Max with both a larger screen and an amplified speaker, for a greater low-end audio experience. Also, product lines such as Chromecast, Google Home, and Google Wifi will now be marketed under the Google Nest brand. In addition, Nest began to deprecate its own internal platforms, announcing the discontinuation of the existing "Works with Nest" program in favor of Google Assistant going forward, and pushing users to migrate themselves from Nest's account system to Google accounts. Google published Nest-specific privacy information outlining a commitment to transparency, not selling personal information, and giving users control of their data. In February 2019, a privacy incident affecting the Google Nest Guard system came about. The controversy stemmed from the fact that Nest Guard, a security device that was part of the Nest Secure system, contained a hidden microphone that was not disclosed in any product specifications. It resulted in a public relations failure. === Partnership with ADT === In August 2020 Google announced intent to invest $450 million in ADT Inc. for a 6.6% stake in the company. The companies intend to integrate Nest devices with ADT's security monitoring services and eventually make them the “cornerstone of ADT’s smart home offering”, according to Nest. Upon the announcement, the shares of ADT doubled in value and hit all-time high of $17.21. === Use with Amazon Alexa === As of mid-2022, Google's newer Nest cameras will now work with Amazon Alexa devices such as Amazon Echo Show, Fire TV, and Fire Tablet to view captured security camera footage. === End of support policies === On October 25, 2025, software support was ended for the 1st and 2nd generation Nest Learning Thermostats. In addition, most of the smart functionality including the Home Away features, notifications, and carbon monoxide sensor became inoperative as they were dependent on connection with Google servers. By mid-November, third-party software solutions became available to restore functionality to affected thermostats. == Products == === Nest Learning Thermostat === The Nest Learning Thermostat is an electronic, programmable, and self-learning Wi-Fi-enabled thermostat that optimizes heating and cooling of homes and businesses to conserve energy. It is based on a machine-learning algorithm: for the first weeks users have to regulate the thermostat in order to provide the reference data set. Nest can then learn people's schedules, at which temperature they are used to and when. Using built-in sensors and phones' locations it can

  Read more →
• 

  Leading the Future

  Leading the Future is an American super PAC network focused on lobbying for policies friendly to the artificial intelligence industry. It was launched in 2025 with over $100 million from industry stakeholders including Andreessen Horowitz, OpenAI President Greg Brockman and Palantir co-founder Joe Lonsdale. The launch was preceded by talks between Collin McCune, head of government affairs at Andreessen Horowitz, and Chris Lehane, chief global affairs officer at OpenAI. Among the members of the network are the American Mission PAC, which supported Chris Gober, and the Think Big PAC, which targeted Alex Bores. Leading the Future is affiliated with the nonprofit Build American AI, which Axios describes as a dark money advocacy "offshoot" operating alongside the super PAC. NBC News states that the network’s efforts are modeled after the pro-cryptocurrency group Fairshake. Leading the Future is led by Zac Moffatt and Josh Vlasto, the latter of whom previously served as an advisor to Fairshake. In response to the creation of Leading the Future, former members of Congress Brad Carson and Chris Stewart co-founded the super PAC network Public First, aiming to counter the group’s influence. In April 2026, an investigation by Model Republic linked Leading the Future to The Wire By Acutus, an automated news website that allegedly used AI agents posing as human journalists to solicit interviews. The site's content was found to closely mirror the PAC's deregulatory policy goals while targeting researchers and advocates skeptical of rapid AI development. In May 2026, Wired revealed that Build American AI used a "dark money" campaign to pay TikTok and Instagram influencers $5,000 per video to promote scripted narratives framing Chinese AI as a "national security threat." According to internal documents and staff at the marketing agency managing the project, the campaign's explicit goal was to "subtly shift public debate" toward the deregulation of AI industries while intentionally avoiding technical discussions regarding AI quality or safety. During the 2026 primary season Leading the Future went on to endorse several candidates in both Democratic and Republican races with several of them going on to win.

  Read more →
• 

  Acoustic model

  An acoustic model is used in automatic speech recognition to represent the relationship between an audio signal and the phonemes or other linguistic units that make up speech. The model is learned from a set of audio recordings and their corresponding transcripts. It is created by taking audio recordings of speech, and their text transcriptions, and using software to create statistical representations of the sounds that make up each word. == Background == Modern speech recognition systems use both an acoustic model and a language model to represent the statistical properties of speech. The acoustic model models the relationship between the audio signal and the phonetic units in the language. The language model is responsible for modeling the word sequences in the language. These two models are combined to get the top-ranked word sequences corresponding to a given audio segment. Most modern speech recognition systems operate on the audio in small chunks known as frames with an approximate duration of 10ms per frame. The raw audio signal from each frame can be transformed by applying the mel-frequency cepstrum. The coefficients from this transformation are commonly known as mel-frequency cepstral coefficients (MFCCs) and are used as an input to the acoustic model along with other features. Recently, the use of convolutional neural networks has led to major improvements in acoustic modeling. == Speech audio characteristics == Audio can be encoded at different sampling rates (i.e. samples per second – the most common being: 8, 16, 32, 44.1, 48, and 96 kHz), and different bits per sample (the most common being: 8-bits, 16-bits, 24-bits or 32-bits). Speech recognition engines work best if the acoustic model they use was trained with speech audio which was recorded at the same sampling rate/bits per sample as the speech being recognized. == Telephony-based speech recognition == The limiting factor for telephony based speech recognition is the bandwidth at which speech can be transmitted. For example, a standard land-line telephone only has a bandwidth of 64 kbit/s at a sampling rate of 8 kHz and 8-bits per sample (8000 samples per second 8-bits per sample = 64000 bit/s). Therefore, for telephony based speech recognition, acoustic models should be trained with 8 kHz/8-bit speech audio files. In the case of voice over IP, the codec determines the sampling rate/bits per sample of speech transmission. Codecs with a higher sampling rate/bits per sample for speech transmission (which improve the sound quality) necessitate acoustic models trained with audio data that matches that sampling rate/bits per sample. == Desktop-based speech recognition == For speech recognition on a standard desktop PC, the limiting factor is the sound card. Most sound cards today can record at sampling rates of between 16–48 kHz of audio, with bit rates of 8- to 16-bits per sample, and playback at up to 96 kHz. As a general rule, a speech recognition engine works better with acoustic models trained with speech audio data recorded at higher sampling rates/bits per sample. But using audio with too high a sampling rate/bits per sample can slow the recognition engine down. A compromise is needed. Thus for desktop speech recognition, the current standard is acoustic models trained with speech audio data recorded at sampling rates of 16 kHz/16 bits per sample.

  Read more →
• 

  Common Crawl

  The Common Crawl Foundation (Common Crawl) is a nonprofit 501(c)(3) organization that crawls the web and freely provides its archives and datasets to the public. Common Crawl was founded by Gil Elbaz. The data had mostly been primarily used by researchers and some startups until the 2020s, when AI companies started training large language models using the data. In November 2025, an investigation by The Atlantic revealed that Common Crawl misled publishers when it claimed it respected paywalls in its scraping and it was not honoring requests from publishers to have their content removed from its databases. == History == Common Crawl was founded in 2007 in San Francisco. It began publishing its crawls in 2011. By 2013, sites like TinEye were building their products off of Common Crawl. The crawl reduces the reliance of companies and researchers on Google, which has the biggest dataset. Common Crawl was designed to have more and fresher data that was more efficient to analyze and utilize than the Wayback Machine created by the Internet Archive. By 2015, 1.8 billion webpages were on the Common Crawl, which started by crawling a list of URLs donated by the search engine Blekko. They use Amazon Web Services, which provides some of its services for free, allowing computing costs to average $2-4000/month. The Common Crawl website listed 30 studies based on Common Crawl data. Before 2023, Common Crawl was not very well known outside of academic researchers who utilize the data. Common Crawl received its first requests to redact information in 2023 and increasingly started seeing its crawler, CCBot, blocked. In 2023, it began receiving significant financial support from AI companies, including Anthropic and OpenAI, each of which donated $250,000. It was also used to train Google DeepMind's large language model Gemini. By April 2023, Common Crawl was capturing 3.1 billion webpages, with an estimated 5% of pages before 2021 containing hate speech or slurs. As of 2024, Common Crawl had been cited in more than 10,000 academic studies. By 2024, The Pile and Common Crawl had been the two main training datasets being used to train AI models. In November 2025, an investigation by technology journalist Alex Reisner for The Atlantic revealed that Common Crawl misled publishers when it claimed it respected paywalls in its scraping and when it said that it was honoring requests from publishers to have their content removed from its databases. It included misleading results in the public search function on its website that showed no entries for websites that had requested their archives be removed, when in fact those sites were still included in its scrapes used by AI companies. As of 2025, Reisner found that CCBot was the most widely-blocked bot by the top 1000 websites. A 2026 article in LWN.net discussed an advantage to services like Common Crawl being that it can limit the scraping costs to websites by allowing companies and researchers to download the data from Common Crawl instead of scraping it themselves. In April 2026, Common Crawl experimentally began to distribute its data through Hugging Face Storage Bucket, in addition to its standard storage on Amazon S3. == Organization == Peter Norvig and Joi Ito have served on the advisory board. Rich Skrenta is the executive director. It has received funding almost exclusively from the Elbaz Family Foundation Trust until 2023 when it started receiving donations from the AI industry. == Refined versions == A number of organizations take raw Common Crawl data and refine it into datasets that exclude edgy content or are otherwise higher-quality for their purposes, such as FineWeb, DCLM and C4. === Colossal Clean Crawled Corpus === Google version of the Common Crawl is called the Colossal Clean Crawled Corpus, or C4 for short. It was constructed for the training of the T5 language model series in 2019. As of 2023, there were some concerns over copyrighted content in the C4 as well as racist content. A 2024 study found that 45% of content was explicitly restricted by websites' terms of service to be used for purposes like AI training by for-profit companies.

  Read more →
• 

  Sarvam AI

  Sarvam AI is an Indian artificial intelligence company headquartered in Bengaluru, Karnataka. Founded in 2023, the company develops large language models (LLMs) and multimodal AI systems with a focus on Indian languages and region-specific use cases. The company has received venture capital backing and has participated in government-supported AI initiatives, including India's sovereign large language model programme under the IndiaAI Mission. == History == Sarvam AI was founded in August 2023 by Vivek Raghavan and Pratyush Kumar, who were previously associated with AI4Bharat at the Indian Institute of Technology Madras. In December 2023, the company announced a combined seed and Series A funding round of approximately US$41 million. The round was led by Lightspeed Venture Partners, with participation from Peak XV Partners and Khosla Ventures. In April 2025, the Ministry of Electronics and Information Technology (MeitY) selected Sarvam AI as one of the companies to develop an indigenous foundational model under the IndiaAI Mission. As part of the initiative, the company received access to government-supported computing infrastructure, including GPUs allocated for model training over a specified period. In February 2026, Sarvam AI introduced two large language models at the AI Impact Summit held at Bharat Mandapam, New Delhi. == Products and technology == Sarvam AI develops language models trained on datasets that include multiple Indian languages and code-mixed text. The company uses mixture-of-experts (MoE) architectures in some of its models. === Foundational language models === On 18 February 2026, the company announced the release of two foundational models: Sarvam-30B – A 30-billion parameter model based on a mixture-of-experts design. According to company disclosures reported by the media, the model activates approximately 1 billion parameters per token and supports a 32,000-token context window. Sarvam-105B – A 105-billion parameter model activating approximately 9 billion parameters per token, with a 128,000-token context window. The model is positioned for complex reasoning and enterprise applications. On 20th February 2026, the company released a beta version of the Sarvam-105B model which is named Indus. It is available on the Apple App Store, Google Play Store and the web. === Speech and vision systems === Sarvam AI has also developed multimodal systems including speech-to-text and vision-language models. Its speech model, referred to as Saaras V3 in company materials, supports multiple Indian languages. The company has also introduced a vision-language model known as Sarvam Vision, intended for document understanding and optical character recognition (OCR) in Indian scripts. === Devices === 'Sarvam Kaze' is an indigenous AI-powered wearable glass that listens, understands, and captures what users see the world through their eyes in real time. The device supports more than 10 Indian languages, enabling voice-based interaction and potentially real-time translation. The company plans to launch the device in May 2026. == Startup support == In March 2026, Sarvam AI launched the Sarvam Startup Program, an initiative providing selected early-stage companies with 6–12 months of API credits scaled to their needs, priority engineering support, and access to production infrastructure for developing multilingual AI applications in areas such as speech, translation, and large language models. == Open-source release == In February 2026, Sarvam AI announced and open-sourced two large language models: Sarvam 30B (30 billion parameters) and Sarvam 105B (105 billion parameters, using a Mixture-of-Experts architecture with 10.3 billion active parameters). Both models were trained from scratch on datasets focused on Indian languages and support advanced reasoning, multilingual tasks, mathematics, and coding. The models are hosted on Hugging Face under the Apache License and are intended for enterprise and developer applications in Indian languages. The models were subsequently released as open source under the Apache License 2.0, with model weights made available on Hugging Face (sarvamai/sarvam-30b and sarvamai/sarvam-105b) and AIKosh in early March 2026. == Government and institutional collaborations == In 2025, Sarvam AI was selected to contribute to India's sovereign AI model initiative under the IndiaAI Mission. The initiative aims to support domestic AI infrastructure and model development. In March 2025, the Unique Identification Authority of India (UIDAI) announced a collaboration with Sarvam AI to integrate AI-based voice interactions and multilingual support into Aadhaar-related services. Sarvam AI has also worked with AI4Bharat and academic institutions on language datasets and speech research projects. == Industry participation == Sarvam AI presented its foundational models at the India AI Impact Summit 2026 in New Delhi. The company has also been listed among Indian members of the AI Alliance, a consortium focused on open-source artificial intelligence initiatives. == List of models ==

  Read more →