Researchers Collected A Dataset of Facial Expressions from Japanese Art

26 February 2020

Researchers Collected A Dataset of Facial Expressions from Japanese Art

A group of researchers has collected and released a new dataset of facial expressions extracted from pre-modern Japanese art. The dataset, called KaoKore contains 5552 RGB image files of size…

New State-of-the-art Self-Supervised Object Tracking Method

22 February 2020

New State-of-the-art Self-Supervised Object Tracking Method

Researchers from the Visual Geometry Group at the University of Oxford have proposed a new self-supervised method for dense tracking which outperforms current state-of-the-art methods in self-supervised visual object tracking.…

GANILLA: GAN Network Trained To Generate Children’s Book Illustrations

19 February 2020

GANILLA: GAN Network Trained To Generate Children’s Book Illustrations

Generative Adversarial Networks have been applied to many domains in image-to-image translation. However, it represents a challenge for GANs to successfully transfer both style and content from one to another…

Microsoft Trained The Largest NLP Model with 17 Billion Parameters

17 February 2020

Microsoft Trained The Largest NLP Model with 17 Billion Parameters

Microsoft announced that it has trained a massive, 17-billion parameter language model that outperforms current state-of-the-art models on many NLP tasks. Turing-NLG, as the new model was called has almost…

Google Proposed A Method for 3D Shape Detection of Transparent Objects

13 February 2020

Google Proposed A Method for 3D Shape Detection of Transparent Objects

In a collaboration with researchers from Synthesis AI and Columbia University, researchers from Google have developed a new method that can detect transparent objects. Detection of transparent objects is a…

Deep Neural Network For Enhancement and Super-resolution of Underwater Images

11 February 2020

Deep Neural Network For Enhancement and Super-resolution of Underwater Images

A group of researchers from the Minessota Robotics Institute has proposed a novel method that uses deep learning for enhancement and super-resolution of underwater images. In their paper, named “Simultaneous…

FAIR Released Billion-scale Dataset for Text Translation

8 February 2020

FAIR Released Billion-scale Dataset for Text Translation

Researchers from Facebook AI Research have open-sourced CCMatrix – a billion-scale text dataset for training translational models. In their blog post, researchers shared details on the dataset, how it was…

Researchers Improve Lane Detection By Generating Night Images Using GAN

7 February 2020

Researchers Improve Lane Detection By Generating Night Images Using GAN

A group of researchers from the Intelligent Navigation Laboratory at Beijing Institute of Technology has proposed a novel method that uses GANs to detect lane markings in low-light conditions. In…

LaserTagger: New State-of-the-art Text Generation Method By Google

4 February 2020

LaserTagger: New State-of-the-art Text Generation Method By Google

Researchers from Google Research, have proposed a new, state-of-the-art method for text generation. In their novel paper, “Encode, Tag, Realize: High-precision Text Editing”, they introduce a novel sequence-to-sequence model that…

Deep Learning Helps Detect Anemia From Eye Images

31 January 2020

Deep Learning Helps Detect Anemia From Eye Images

Researchers from Google Health have developed a method that uses deep learning to detect anemia only from the eye. In their recent paper published in Nature Biomedical Engineering, researchers describe…

FAIR Open-Sourced Polygames – A New Environment for Training Agents via Self-play

30 January 2020

FAIR Open-Sourced Polygames – A New Environment for Training Agents via Self-play

Facebook AI has announced the release of Polygames – a new open-source environment for training AI agents to master strategy games through self-play. The goal of Polygames is to provide…

Meena: Google’s New State-of-the-art Chatbot

29 January 2020

Meena: Google’s New State-of-the-art Chatbot

Researchers from Google Brain have presented Meena – the new state-of-the-art chatbot that can chat almost about anything. In their new paper, named “Towards a Human-like Open-domain Chatbot”, researchers describe…

Attentive Graph Neural Networks: New Method For Video Object Segmentation

27 January 2020

Attentive Graph Neural Networks: New Method For Video Object Segmentation

Researchers from the Inception Institute of Artificial Intelligence, UAE and Indiana University have proposed a novel state-of-the-art method for video object segmentation based on graph neural networks. In their work,…

Neural Network Segments Objects In Videos Without Using Labeled Data

25 January 2020

Neural Network Segments Objects In Videos Without Using Labeled Data

In a joint work, researchers from several universities have developed a novel method for unsupervised video object segmentation that significantly outperforms other existing methods at this task. The new deep…

Google Dataset Search Comes Out of Beta Version

24 January 2020

Google Dataset Search Comes Out of Beta Version

Today, Google has announced the official release of Google Dataset Search. The new feature allows for discovering millions of datasets available on the web. Engineers at Google have indexed more…

FixMatch: New State-of-the-art Method In Semi-Supervised Learning

23 January 2020

FixMatch: New State-of-the-art Method In Semi-Supervised Learning

Researchers from Google Research have proposed a new state-of-the-art method in semi-supervised image classification. The problem of image classification has been successfully tackled in the past several years thanks to…

Advbox: Baidu Released Toolbox for Generating Adversarial Examples

21 January 2020

Advbox: Baidu Released Toolbox for Generating Adversarial Examples

In the past few years, researchers have discovered numerous ways on how to fool neural networks. The input samples that can do that were called adversarial examples and they were…

CenterMask: New State-of-the-art Model In Instance Segmentation

20 January 2020

CenterMask: New State-of-the-art Model In Instance Segmentation

Researchers from the Electronics and Telecommunications Research Insitute of Korea have proposed a new state-of-the-art method for instance segmentation. The novel method called CenterMask is based on a fully-convolutional one-stage…

Everybody’s Talkin’: New Method Generates Realistic Videos Given Audio Input

19 January 2020

Everybody’s Talkin’: New Method Generates Realistic Videos Given Audio Input

A group of researchers from SenseTime Research, NLPR, and Nanyang Technological University has proposed a novel method that generates a video of a person talking, given audio signal and a…

HybridPose: New State-of-the-art Method for 6D Pose Estimation

16 January 2020

HybridPose: New State-of-the-art Method for 6D Pose Estimation

Group of researchers from the University of Texas, Austin has developed a novel 6D pose estimation method that significantly outperforms existing methods. Their new approach, called HybridPose, predicts an intermediate…

Google Uses Convolutional Neural Networks To Precisely Forecast Precipitation

15 January 2020

Google Uses Convolutional Neural Networks To Precisely Forecast Precipitation

In a recent paper, researchers from Google AI Research propose a method for precipitation forecasting using deep learning and convolutional neural networks. Their approach, different than traditional weather models, does…