Trending Research

Towards VQA Models That Can Read

CVPR 2019 facebookresearch/pythia

We show that LoRRA outperforms existing state-of-the-art VQA models on our TextVQA dataset.

VISUAL QUESTION ANSWERING

1,025
1.73 stars / hour

Pythia v0.1: the Winning Entry to the VQA Challenge 2018

26 Jul 2018facebookresearch/pythia

We demonstrate that by making subtle but important changes to the model architecture and the learning rate schedule, fine-tuning image features, and adding data augmentation, we can significantly improve the performance of the up-down model on VQA v2. 0 dataset -- from 65. 67% to 70. 22%.

VISUAL QUESTION ANSWERING

1,025
1.73 stars / hour

MixMatch: A Holistic Approach to Semi-Supervised Learning

6 May 2019google-research/mixmatch

Semi-supervised learning has proven to be a powerful paradigm for leveraging unlabeled data to mitigate the reliance on large labeled datasets.

SEMI-SUPERVISED IMAGE CLASSIFICATION

255
1.00 stars / hour

ERNIE: Enhanced Language Representation with Informative Entities

17 May 2019thunlp/ERNIE

Neural language representation models such as BERT pre-trained on large-scale corpora can well capture rich semantic patterns from plain text, and be fine-tuned to consistently improve the performance of various NLP tasks.

KNOWLEDGE GRAPHS

110
0.94 stars / hour

Multinomial Distribution Learning for Effective Neural Architecture Search

18 May 2019tanglang96/MDENAS

Therefore, NAS can be transformed to a multinomial distribution learning problem, i. e., the distribution is optimized to have high expectation of the performance.

ARCHITECTURE SEARCH

24
0.42 stars / hour

Star-Transformer

25 Feb 2019fastnlp/fastNLP

Although Transformer has achieved great successes on many NLP tasks, its heavy structure with fully-connected attention connections leads to dependencies on large training data.

NATURAL LANGUAGE INFERENCE TEXT CLASSIFICATION

482
0.39 stars / hour

Making Convolutional Networks Shift-Invariant Again

25 Apr 2019adobe/antialiased-cnns

The well-known signal processing fix is anti-aliasing by low-pass filtering before downsampling.

CONDITIONAL IMAGE GENERATION IMAGE CLASSIFICATION

36
0.38 stars / hour
44
0.36 stars / hour

Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context

ICLR 2019 huggingface/pytorch-pretrained-BERT

Transformer networks have a potential of learning longer-term dependency, but are limited by a fixed-length context in the setting of language modeling.

LANGUAGE MODELLING

6,306
0.32 stars / hour