Google powers YouTube with its own Video-Transcoding Chips

Google AI chip

YouTube is the world’s most popular video-sharing platform. Keeping it going was once thought to be difficult before Google purchased it in 2006. Since then, Google has fought hard to keep the site’s costs down, often reinventing Internet technology and copyright to do so. The primary infrastructure issue that YouTube must address for end-users today is delivering content optimized for your platform and bandwidth while ensuring efficiency.

Read More

Google announces the general availability of Vertex AI to expedite the development and maintenance of Artificial Intelligence (AI) Models

Google Vertex

Data scientists, every day, face manually stitching together Machine Learning point solutions and finding anomalies, resulting in a lag in model creation and experimentation, hence reducing the production level. To address these issues, Google announced Vertex AI, a managed machine learning platform meant to expedite the development and maintenance of artificial intelligence (AI) models to be generally available.

Read More

Project Starline by Google promises to usher in a new era of Video Conferencing with 3D Display

Project Starline - 3D display

Human beings are social animals who like to connect with their counterparts in one way or another. While we have reached significant milestones in the arena of technology, communication remains a considerable roadblock. An entire plethora of video conferencing platforms like Zoom, Google Meet, Teams, among significant others, have been made available to the masses and are in constant use, especially due to the pandemic. However, all of them are far behind the actual face-to-face talking experience, thereby offering only limited connectivity.

Read More

[Paper] Facebook AI releases Dynaboard: A New Evaluation platform for NLP Models

Last year, Facebook AI released Dynabench, a platform that radically rethinks benchmarking in AI, starting with natural language processing (NLP) models. Going forward, they have now announced a new evaluation-as-a-service platform for comprehensive, standardized evaluations of NLP models called Dynaboard. Dynaboard can perform apples-to-apples comparisons dynamically without common issues from bugs in evaluation code, inconsistencies in filtering test data, backward compatibility, accessibility, and several other reproducibility issues.

Read More

This Washington-Based AI Startup offers an Acceleration Platform

Octo ML Acceleration platform

OctoML is a Washington-based startup that offers an acceleration platform for deploying machine learning models and algorithms on the hardware. This platform primarily helps the engineering teams deploy the machine learning models seamlessly and with increased accuracy. The platform is built on an open-source Apache TVM compiler framework project. In the recent Series B funding rounds, OctoML raised $28 million, which takes the company’s total capital to around $47 million.

Read More

[Paper Summary] Researchers from NVIDIA, Stanford University and Microsoft Research propose Efficient Trillion-Parameter Language Model Training on GPU Clusters

In a paper by NVIDIA, Stanford University, and Microsoft Research, a research team has proposed a new parallelization schedule that improves throughput by more than 10 percent with a comparable memory footprint. The paper demonstrated that such strategies could be composed to achieve high aggregate throughput when training large models with nearly a trillion parameters. 

Read More

[Paper Summary] Google AI proposes a Machine Learning Algorithm for teaching Agents to solve new tasks

Google algorithm graph

Google AI suggests an alternative, example-based control, which aims at teaching agents how to solve new tasks by providing examples of success. This is termed as recursive classification of examples (RCE). It does not rely on formulated reward functions, distance functions, or features. It instead just uses the examples of success.

Read More

Researchers from DeepMind and Alberta University propose policy-guided Heuristic search Algorithm

DeepMind’s AlphaGo and its successors previously demonstrated that the policy and heuristic function is formulated upon the PUCT (Polynomial Upper Confidence Trees) search algorithm. This algorithm can be quite effective for guiding search in adversarial games. However, PUCT is computationally inefficient and lacks guarantees on its search effort. Though other methods such as LevinTS provide guarantees on search steps, they do not use a heuristic function.

Read More

Google AI introduces a new system for Open-Domain Long-Form Question Answering (LFQA)

Attention

Open-domain long-on answering (LFQA) form questions a fundamental challenge in natural language processing (NLP) that involves retrieving documents relevant to a given query and using them to generate a detailed paragraph-length answer. 

Read More

[Paper Summary] Researchers at the University Of Genoa and AWS analyze techniques to Make Machine Learning (ML) Models fairer

AI fair or unfair

This research on algorithmic fairness provides three main approaches, i.e., pre-processing data, post-processing an already learned ML model, and in-processing, which consists of enforcing fairness notions by imposing specific statistical constraints on the learning phase of the model.

Read More

Google AI, DeepMind and the University of Toronto introduce DreamerV2

dreamer-v2

It is the first Reinforcement Learning (RL) agent based on the world model to attain human-level success on the Atari benchmark. It includes the second generation of the Dreamer agent who learns behaviors entirely within a world model’s latent space […]

Read More

Georgia Tech and Facebook AI researchers devise a new Tensor…

A recent study conducted jointly by the Georgia Institute and Facebook AI researchers has opened the door to a new method called TT-Rec (Tensor-Train for DLRM). If employed successfully, this method would be a leap forward in the arena of deep learning as it will significantly reduce the size of the Deep Learning Recommendation Models […]

Read More

Allen Institute For AI (AI2) Launches The 2.7.0 Version of AI2-THOR That Enables Users To Reduce Their Training Time Dramatically

AI2-THOR

Allen Institute for AI (AI2) has recently announced the 2.7.0 release of AI2-THOR. AI2-THOR is an open-source interactive environment for training and testing embodied AI. The 2.7.0 version of AI2-THOR contains several performance enhancements that can provide dramatic training time reductions. The new version introduces improvements to the IPC system between Unity/Python and serialization/deserialization format. It includes new actions that are much better to control the metadata. 

Read More

Embold: Static Code Analyzer uses AI to help Developers analyze and improve their code

Embold is a simple but efficient AI-based static code analyzer that can help developers analyze and improve their code. The feature that truly makes it stand apart is its ability to analyze source code across four dimensions: code issues, design issues, metrics and duplication, and surface issues that impact stability, robustness, security, and maintainability.

Read More

Intel to acquire SigOpt, an AI Hyperparameter Optimization Platform

Intel has confirmed that it is buying SigOpt Inc., an artificial intelligence startup developing software platforms to optimize AI models. Several private firms and research groups such as OpenAI use these software platforms to boost their AI models’ performance.

Read More

Microsoft introduces Lobe: A free Machine Learning application that allows you to create AI Models without coding

Microsoft has released Lobe, a free desktop application that lets Windows and Mac users create customized AI models without writing any code. Several customers are already using the app for tracking tourist activity around coral reefs, the company said.

Read More