Trending Research

Kaolin: A PyTorch Library for Accelerating 3D Deep Learning Research

12 Nov 2019NVIDIAGameWorks/kaolin

We present Kaolin, a PyTorch library aiming to accelerate 3D deep learning research.

478
10.96 stars / hour

Contrastive Representation Distillation

23 Oct 2019HobbitLong/RepDistiller

We demonstrate that this objective ignores important structural knowledge of the teacher network.

MODEL COMPRESSION TRANSFER LEARNING

262
1.34 stars / hour

Global Sparse Momentum SGD for Pruning Very Deep Neural Networks

NeurIPS 2019 ShawnDing1994/ACNet

Deep Neural Network (DNN) is powerful but computationally expensive and memory intensive, thus impeding its practical usage on resource-constrained front-end devices.

MODEL COMPRESSION

121
1.30 stars / hour

ACNet: Strengthening the Kernel Skeletons for Powerful CNN via Asymmetric Convolution Blocks

ICCV 2019 ShawnDing1994/ACNet

We propose Asymmetric Convolution Block (ACB), an architecture-neutral structure as a CNN building block, which uses 1D asymmetric convolutions to strengthen the square convolution kernels.

121
1.30 stars / hour

The Measure of Intelligence

5 Nov 2019fchollet/ARC

To make deliberate progress towards more intelligent and more human-like artificial systems, we need to be following an appropriate feedback signal: we need to be able to define and evaluate intelligence in a way that enables comparisons between two systems, as well as comparisons with humans.

431
1.18 stars / hour

Language Models are Unsupervised Multitask Learners

Preprint 2019 openai/gpt-2

Natural language processing tasks, such as question answering, machine translation, reading comprehension, and summarization, are typically approached with supervised learning on taskspecific datasets.

 SOTA for Language Modelling on Text8 (using extra training data)

COMMON SENSE REASONING DOCUMENT SUMMARIZATION LANGUAGE MODELLING MACHINE TRANSLATION QUESTION ANSWERING READING COMPREHENSION TEXT GENERATION

9,653
0.82 stars / hour

CenterFace: Joint Face Detection and Alignment Using Face as Point

9 Nov 2019Star-Clouds/CenterFace

Face detection and alignment in unconstrained environment is always deployed on edge devices which have limited memory storage and low computing power.

FACE DETECTION

363
0.73 stars / hour

Scale-Aware Trident Networks for Object Detection

ICCV 2019 facebookresearch/detectron2

In this work, we first present a controlled experiment to investigate the effect of receptive fields for scale variation in object detection.

OBJECT DETECTION

5,472
0.66 stars / hour

DialoGPT: Large-Scale Generative Pre-training for Conversational Response Generation

1 Nov 2019microsoft/DialoGPT

We present a large, tunable neural conversational response generation model, DialoGPT (dialogue generative pre-trained transformer).

CONVERSATIONAL RESPONSE GENERATION

269
0.60 stars / hour