Facebook AI’s Voice Separation Model That Distinguishes Multiple Speakers Simultaneously

At the time of isolation and social distancing due to the pandemic, there are many changes in everyone's life. One among them...

Penrose: A system to convert complex mathematical equations into simple visual diagrams

Researchers from CMU and Technion developed Penrose. It is named Penrose, after the renowned mathematician and physicist Roger Penrose. The unavailability of...

GPT-3: A New Breakthrough in Language Generator

OpenAI has come up with a language generator GPT-3, which is a successor of GPT-2. This newly developed AI was put forward...

Libra: A Python Tool That Automates Machine Learning Process in a Few Lines of...

This python library 'Libra' automates the end-to-end machine learning process with just one line of code. Libra is built for both software...

Continuous Machine Learning (CML): An Open-Source CI/CD library for Machine Learning

Totally, open-source, and made by the same development team that created the popular DVC (data version control) library, CML (Continous Machine Learning) library is...

Top Artificial Intelligence Books to Read in 2020

Artificial Intelligence: A Modern Approach by Stuart Russell and Peter Norvig: A Modern Approach, 3e offers the most comprehensive, up-to-date introduction to the theory...

Exclusive Talk with Ashutosh Saxena, Founder & CEO at Caspar.AI

Ashutosh Saxena is the Founder and CEO of Caspar.AI, that builds AI-enabled homes for senior and luxury housing. Dr....

ABBYY Open-Sources NeoML, A Machine Learning Framework For Both Deep Learning And Traditional Algorithms

ABBYY recently announced the launch of NeoML, an open-source library for building, training, and deploying machine learning models. NeoML is a cross-platform framework...

Neuropod: Uber ATG’s Open Source Deep Learning Inference Engine

Uber Advanced Technologies Group (ATG) recently open-sourced Neuropod, a framework that provides a uniform interface to run deep learning models from multiple...

Hummingbird: A library for compiling trained traditional machine learning models into tensor computations

This is really a cool work out of Microsoft research called hummingbird. You can convert traditional machine learning models to tensor computations...