Stratified normalization: a new method for improving the performance of neural networks

16 April 2021

Stratified normalization: a new method for improving the performance of neural networks

The new stratified normalization algorithm significantly increases the accuracy of emotion recognition based on the electroencephalogram (EEG) signal. Feature normalization per participant and session – the key idea of the…

Fairness Flow: AI Bias Assessment Tool

7 April 2021
Fairness flow

Fairness Flow: AI Bias Assessment Tool

Fairness Flow is a set of diagnostic tools that helps you compare how fair models and label markers perform for specific user groups. Why is it needed Facebook AI specialists…

The neural network generates images with clothes try on

14 March 2021

The neural network generates images with clothes try on

PF-AFN is a neural network that generates images of people trying on different kinds of clothes. The model accepts an image of a person and an image of a garment…

SEER: a self-supervised neural network with a billion parameters from FAIR

9 March 2021

SEER: a self-supervised neural network with a billion parameters from FAIR

SEER is FAIR’s self-supervised billion-parameter neural network for computer vision applications. The model pre-trained on the Instagram pictures can be further trained on your tasks. The developers have published the…

Robot trained to manipulate tissue at University of California

5 March 2021

Robot trained to manipulate tissue at University of California

Researchers from the University of California and the Honda Research Institute trained a robot to fold fabric. The algorithm is based on a framework for teaching visual dynamics of objects…

GraphGallery: a library for graph neural networks on PyTorch and TensorFlow

2 March 2021

GraphGallery: a library for graph neural networks on PyTorch and TensorFlow

GraphGallery is a library for training and testing graph neural networks. GraphGallery implements adversarial attacks on graph neural networks. The library is compatible with PyTorch, TensorFlow 2.x, Pytorch Geometric (PyG),…

Google presented the framework for ML-model architecture automatic search

23 February 2021

Google presented the framework for ML-model architecture automatic search

Model search (MS) is a library that uses ML model architecture automatic search algorithms. The developers claim that the framework scales in cases when the state search space appears large.…

Robot manage objects from video tutorials using RL

19 February 2021

Robot manage objects from video tutorials using RL

In FAIR, the RL-agent was trained to manage objects using video tutorials. Standard RL algorithms are trained to a problem iteratively through learning from errors. The proposed algorithm learns a…

MeInGame neural network generates game character from a face image

15 February 2021

MeInGame neural network generates game character from a face image

MeInGame is a neural network model that generates a character in the game from one face image. The neural network predicts the shape of the face and its texture. The…

JigsawGAN: Generative neural network model solves jigsaw puzzles

11 February 2021

JigsawGAN: Generative neural network model solves jigsaw puzzles

JigsawGAN is a self-supervised generative neural network model that has been trained on a puzzle-solving task. The model accepts chaotically located parts of the image as input and outputs the…

FaceX-Zoo: PyTorch library for face recognition in images

29 January 2021

FaceX-Zoo: PyTorch library for face recognition in images

FaceX-Zoo is an open-source library on PyTorch for recognizing faces in images. The library provides a module for training models with different configurations of error functionality and basic architecture. In…

RxR: Google Released New Dataset and Challenge On Robot Navigation Using Language

22 January 2021

RxR: Google Released New Dataset and Challenge On Robot Navigation Using Language

A group of researchers from Google has announced the release of a new dataset and benchmark for navigation instruction called Room-across-Room (RxR). Robotic navigation is one of the major challenges…

New AI System Can Predict If a COVID Patient Will Need Intensive Care

21 January 2021

New AI System Can Predict If a COVID Patient Will Need Intensive Care

Researchers from Facebook AI have open-sourced a method and a system that can predict if a Covid patient will need intensive care, using chest x-ray images and self-supervised learning. In…

PaddleSeg: A New Toolkit for Efficient Image Segmentation

20 January 2021

PaddleSeg: A New Toolkit for Efficient Image Segmentation

Researchers from Baidu Inc. have announced the release of a novel image segmentation toolkit called PaddleSeg. The development kit was built in order to help researchers and engineers working in…

Switch Transformer: Google’s New Language Model Features Trillion Parameters

19 January 2021

Switch Transformer: Google’s New Language Model Features Trillion Parameters

There’s been a lot of hype around OpenAI’s powerful GPT-3 model, which proved to be able to spin novel super-realistic human-like articles but also tackle many different NLP tasks using…

Researchers Re-labeled ImageNet Introducing Multi-labels and Localized Annotations

17 January 2021

Researchers Re-labeled ImageNet Introducing Multi-labels and Localized Annotations

A large portion of the research experiments in Computer Vision in the past decade were using the ImageNet dataset. Over the years it has become the default and standard benchmark…

Pr-VIPE: New Method Successfully Recognizes 3D Poses in 2D Images

15 January 2021

Pr-VIPE: New Method Successfully Recognizes 3D Poses in 2D Images

Researchers from Google AI have presented a new method for recognizing poses and pose similarity in images and videos. Images and videos contain 2D information about some portion of the…

Novel Neural Network Generates Segmentation Masks from Bounding Boxes in Videos

13 January 2021

Novel Neural Network Generates Segmentation Masks from Bounding Boxes in Videos

A group of researchers from ETH University has proposed a novel method for object semantic segmentation that exploits video data and spatio-temporal consistencies. Arguing that current approaches for object segmentation…

Researchers Design a Plain Simple Network that Achieves Over 80% Accuracy on ImageNet

13 January 2021

Researchers Design a Plain Simple Network that Achieves Over 80% Accuracy on ImageNet

In a joint project, researchers from several universities in China and UK have proposed a new powerful neural network which has a VGG-like plain and simple architecture using only a…

DALL-E: OpenAI’s New 12 Billion Parameter Model

7 January 2021

DALL-E: OpenAI’s New 12 Billion Parameter Model

Researchers from OpenAI have presented their new model that can generate realistic images given text descriptions. The model, called DALL-E is based on OpenAI’s powerful GPT-3 model and can generate…

Lambda Networks: New State-of-the-art Architecture for Image Recognition

2 January 2021

Lambda Networks: New State-of-the-art Architecture for Image Recognition

LambdaNetworks – researchers propose a new type of deep neural networks that are computationally efficient while maintaining on-par performance with existing classification models. Based on an interesting idea, Lambda Networks…