Deepfake Videos: GAN Sythesizes a Video From a Single Photo

25 May 2019

Deepfake Videos: GAN Sythesizes a Video From a Single Photo

Researchers from Samsung AI and Skolkovo Institute of Science and Technology have produced a system that can create realistic fake videos of a person talking, given only a few images…

PyTorch’s torchvision 0.3 Comes With Segmentation and Detection Models, New Datasets and More

25 May 2019

PyTorch’s torchvision 0.3 Comes With Segmentation and Detection Models, New Datasets and More

The new release 0.3 of PyTorch’s torchvision library brings several new features and improvements. The newest version of torchvision includes models for semantic segmentation, instance segmentation, object detection, person keypoint…

“Moving Camera, Moving People”: Novel Method for Depth Estimation from Google AI

24 May 2019

“Moving Camera, Moving People”: Novel Method for Depth Estimation from Google AI

Researchers from Google AI have proposed a new method for depth estimation of moving people. The new method is capable of handling videos where both the people and the camera…

Facebook Open-sourced Pythia: Deep Learning Framework for Image and Language Models

24 May 2019

Facebook Open-sourced Pythia: Deep Learning Framework for Image and Language Models

Facebook AI Research (FAIR) has released Phytia – an open-source framework for multimodal learning. Built on top of PyTorch, Pythia supports multitask learning for the vision and language domains. Multimodal…

PaperRobot: Automatic Research Assistant That Can Generate Drafts of Scientific Ideas

22 May 2019

PaperRobot: Automatic Research Assistant That Can Generate Drafts of Scientific Ideas

In a paper published today on Arxiv, researchers propose a method that generates drafts of scientific ideas and acts as an automatic research assistant. PaperRobot, as it was named, is…

Artificial intelligence Diagnoses Lung Cancer Better Than Human Experts

22 May 2019

Artificial intelligence Diagnoses Lung Cancer Better Than Human Experts

A group of researchers from Google AI, Stanford University School of Medicine, Northwestern Medicine and NYU Langone Health published their work on early lung cancer diagnosis with machine learning. The…

“MixMatch”: New Semi-Supervised Learning Approach Achieves State-of-the-art Results

20 May 2019

“MixMatch”: New Semi-Supervised Learning Approach Achieves State-of-the-art Results

Researchers from Google Research have proposed a novel approach to semi-supervised learning that achieves state-of-the-art results in many datasets and with different labeled data amounts. In semi-supervised learning, methods use…

Samsung AI Released A New Method For Human Pose Tracking

18 May 2019

Samsung AI Released A New Method For Human Pose Tracking

Researchers from Samsung AI and the Skolkovo Institute of Science and Technology have published a paper where they propose a new state-of-the-art method for human pose tracking. In the paper,…

Self-supervised Learning Method For Video Flow Correspondence

17 May 2019

Self-supervised Learning Method For Video Flow Correspondence

Researchers from the Department of Computer Science and Visual Geometry Group at the University of Oxford have proposed a new self-supervised learning method for video correspondence flow. The task of…

“Translatotron”: Google Releases Translation Model That Mimics Human Voice

16 May 2019

“Translatotron”: Google Releases Translation Model That Mimics Human Voice

Researchers from Google AI have proposed a new speech-to-speech translation model and a new experimental system that shows faster inference and improved quality of the translation. The new translation model…

Google’s AutoML System Competed Against Humans In A Kaggle Competition

15 May 2019

Google’s AutoML System Competed Against Humans In A Kaggle Competition

Google’s AutoML system for tabular data took part of Kaggle Days Hackathon and entered a machine learning competition for the first time. AutoML (as it is nicknamed) was competing against…

Computer Graphics Meets Deep Learning with TensorFlow Graphics

14 May 2019

Computer Graphics Meets Deep Learning with TensorFlow Graphics

Google announced the first release of TensorFlow Graphics – differentiable graphics layers for TensorFlow which connect deep learning with computer graphics. In the past years, a lot of research within…

Google AI Announced Open Images v5 Dataset

11 May 2019

Google AI Announced Open Images v5 Dataset

Google AI announced Open Images v5 – a new version of Google’s large Open Images dataset which introduces segmentation masks to the set of annotations. The new dataset contains segmentation…

New Neural Network Achieves State-of-the-art Performance on Person Reidentification

8 May 2019

New Neural Network Achieves State-of-the-art Performance on Person Reidentification

Researchers from the University of Surrey, the Queen Mary University of London and Samsung AI have proposed a novel neural network model that achieves state-of-the-art performance in person re-identification. Person…

Facebook Launches PyTorch 1.1 with Tensorflow’s TensorBoard Support

5 May 2019

Facebook Launches PyTorch 1.1 with Tensorflow’s TensorBoard Support

On the second day of Facebook’s annual developer conference F8, the company announced the arrival of PyTorch 1.1. The improvement is a big milestone for PyTorch and includes new developer…

Google Released World’s Largest Landmarks Dataset With 5 Million Images

4 May 2019

Google Released World’s Largest Landmarks Dataset With 5 Million Images

Google AI has announced the release of the world’s largest landmark recognition dataset called Google-Landmarks-v2. The dataset is the extension or the second version of the Google-Landmarks dataset released last…

FaceStyleGAN: GAN Network Generates Style Portraits in Snapchat

4 May 2019

FaceStyleGAN: GAN Network Generates Style Portraits in Snapchat

Researchers from the Czech Technical University and Snap Inc. have proposed a novel method that given a portrait image and a style exemplar generates a portrait of a desired style.…

Introducing “UDA”: Unsupervised Data Augmentation Method That Achieves State-of-the-art Results

1 May 2019

Introducing “UDA”: Unsupervised Data Augmentation Method That Achieves State-of-the-art Results

Researchers from Google Brain and Carnegie Mellon University have proposed a new data augmentation method that achieves state-of-the-art results on IMDB Dataset (with just 20 labeled examples), SSL Cifar10 and…

OpenAI’s MuseNet AI Can Generate Novel 4-Minute Music Compositions

29 April 2019

OpenAI’s MuseNet AI Can Generate Novel 4-Minute Music Compositions

OpenAI has released MuseNet – a deep neural network that creates musical compositions with 10 different instruments. The neural network model is based on the general unsupervised method as the…

“Segmenting The Future”: New Method Predicts Segmentation Masks From Past Video Frames

26 April 2019

“Segmenting The Future”: New Method Predicts Segmentation Masks From Past Video Frames

A novel method proposed by researchers at Stanford University encodes RGB frames from videos and predicts future semantic segmentation. In the past, researchers have tackled a wide variety of prediction…

OpenAI Released Sparse Transformer Models

25 April 2019

OpenAI Released Sparse Transformer Models

Open AI released Sparse Transformers – deep neural network models which set new records in predicting what comes next in a sequence. Researchers from OpenAI proposed an improvement to the…