Cylindrical and Asymmetrical 3D Convolution Networks: State-of-the-art in Lidar Segmentation

25 November 2020

Cylindrical and Asymmetrical 3D Convolution Networks: State-of-the-art in Lidar Segmentation

A group of researchers from several Chinese universities has developed a novel state-of-the-art method for LIDAR semantic segmentation using asymmetrical 3D convolutional networks. In their paper, “Cylindrical and Asymmetrical 3D…

CompressAI: New Library for Deep-Learning-Based Compression

17 November 2020

CompressAI: New Library for Deep-Learning-Based Compression

CompressAI – a new PyTorch library for deep learning compression research was released this week. The new library aims to provide researchers with a flexible tool for doing research and…

Novel Method Successfully Removes Objects from Videos

17 November 2020

Novel Method Successfully Removes Objects from Videos

Researchers from Virginia Tech and Facebook have developed a video completion method that can smoothly remove objects from videos. In their paper, “Flow-edge Guided Video Completion” the group describes their…

Neural Network Synthesizes Images Given Text Description

10 November 2020

Neural Network Synthesizes Images Given Text Description

A group of researchers from Google Research has proposed a novel method for text-to-image generation that outperforms existing direct text-to-image generation models. In their work, researchers designed a sequential neural…

Google AI Released A New 3D Object Detection Pipeline and Dataset

10 November 2020

Google AI Released A New 3D Object Detection Pipeline and Dataset

As part of a larger project aimed to improve and bring accurate 3D object detection on mobile devices, researchers from Google announced the release of large-scale video dataset with 3D…

AI Identifies Deepfakes Using Heartbeat Detection

5 November 2020

AI Identifies Deepfakes Using Heartbeat Detection

Deepfakes have been a popular topic in the past few years. The fake portrait videos generated using generative models have been used widely for many different use cases from celebrity…

Researchers Used Deep Learning to Detect COVID-19 Disease

26 October 2020

Researchers Used Deep Learning to Detect COVID-19 Disease

A group of researchers from Alanya Alaaddin Keykubat University has used deep learning and convolutional neural network to detect Coronavirus COVID-19 disease. In their paper published at Physiol Genomics 2020,…

Vision Transformers: Transformers Work Well in Computer Vision Too

25 October 2020

Vision Transformers: Transformers Work Well in Computer Vision Too

The Transformer architecture or Transformer models have been one of the largest breakthroughs in AI in the past decade. One of the most powerful models, namely GPT-3 is based on…

New Neural Network Model Puts All Objects Into Focus

24 October 2020

New Neural Network Model Puts All Objects Into Focus

In images, almost always the main objects within the depth-of-field are in focus and all other objects including the background are blurred, at least to some extent. This happens due…

Researchers Find Out The Role Of Individual Units of a Deep Network

23 October 2020

Researchers Find Out The Role Of Individual Units of a Deep Network

A group of researchers has proposed and developed neural network dissection methods, and have used them in order to explore what role do individual units of a deep neural network…

Neural Network Generates 3D Human Model Out Of A Selfie Photo

23 October 2020

Neural Network Generates 3D Human Model Out Of A Selfie Photo

A group of researchers from Tencent AI Labs, has developed and open-sourced a new method for high-fidelity 3D model reconstruction of humans by using RGB-D selfie images. The method works…

Facebook Announced “The Open Catalyst Project”: Using AI For Renewable Energy

15 October 2020

Facebook Announced “The Open Catalyst Project”: Using AI For Renewable Energy

Facebook has announced a collaboration with Carnegie Mellon University on a new project called the Open Catalyst Project, whose goal is to leverage advances in AI towards the discovery of…

Torch-Points3D: New Modular Framework For 3D Deep Learning

14 October 2020

Torch-Points3D: New Modular Framework For 3D Deep Learning

A group of researchers has joined forces to develop an open-source modular framework for facilitating work on deep neural networks for 3D point cloud data – Torch-Points3D. The new framework…

NVIDIA Researchers Improved StyleGAN2 Via New Augmentation Method

13 October 2020

NVIDIA Researchers Improved StyleGAN2 Via New Augmentation Method

GAN networks achieved great success in the past few years. Among them, StyleGAN2 stood out as one of the most powerful generative models, which was able to model high-resolution images…

Researchers Pre-train Point Cloud Models by Point Cloud Completion

11 October 2020

Researchers Pre-train Point Cloud Models by Point Cloud Completion

In a recent paper, a group of researchers from Oxford, Cambridge, UC Berkeley and UCL has proposed a new pre-training method for point cloud models that relies on point cloud…

NVIDIA Released Imaginaire: New Library for GAN Models

7 October 2020

NVIDIA Released Imaginaire: New Library for GAN Models

Researchers from NVIDIA have announced the release of Imaginaire – a new PyTorch library for generative modeling, which hosts optimized implementations of various GAN-based image and video synthesis methods. Imaginaire…

Facebook Researchers Used Reinforcement Learning to Accelerate MRI Scanning

5 October 2020

Facebook Researchers Used Reinforcement Learning to Accelerate MRI Scanning

A group of researchers from Facebook AI Research and McGill University has released a new state-of-the-art method for active MRI acquisition that uses Reinforcement Learning. In their novel paper, “Active…

Google Researchers Released New Instance-level Recognition Model

28 September 2020

Google Researchers Released New Instance-level Recognition Model

In a novel paper, “Unifying Deep Local and Global Features for Image Search”, researchers from Google proposed a new method for image searching that unifies local and global features, enabling…

Microsoft Researchers Have Found a Way to Compare Any Two Datasets

26 September 2020

Microsoft Researchers Have Found a Way to Compare Any Two Datasets

In recent work, researchers from Microsoft have proposed a new method for measuring dataset similarity based on optimal transport (OT) techniques. Measuring dataset similarity is a non-trivial problem that has…

Deep Neural Network Restores Severely Damaged Old Photos

26 September 2020

Deep Neural Network Restores Severely Damaged Old Photos

A group of researchers from several universities have joined forces on developing a new method for old image restoration using deep learning. In their paper, “Old Photo Restoration via Deep…

PP-OCR: Novel Ultra-lightweight Model for Optical Character Recognition

24 September 2020

PP-OCR: Novel Ultra-lightweight Model for Optical Character Recognition

A group of researchers from Baidu has proposed and developed a novel ultra-lightweight optical character recognition system (OCR). Their, so-called PP-OCR is based on lightweight deep neural networks with several…