Heterogeneous Graph Neural Networks (HGNNs) have gained significant
popu...
Node classification is a substantial problem in graph-based fraud detect...
Sparsely gated Mixture-of-Expert (MoE) has demonstrated its effectivenes...
Arbitrary-oriented object detection is a fundamental task in visual scen...
The accuracy of AlphaFold2, a frontier end-to-end structure prediction
s...
A large amount of document data exists in unstructured form such as raw
...
Recent efforts of multimodal Transformers have improved Visually Rich
Do...
Pre-training models are an important tool in Natural Language Processing...
Although more layers and more parameters generally improve the accuracy ...
Accurate protein structure prediction can significantly accelerate the
d...
Optical character recognition (OCR) technology has been widely used in
v...
PaddleSpeech is an open-source all-in-one speech toolkit. It aims at
fac...
With the increasing diversity of ML infrastructures nowadays, distribute...
The ever-growing model size and scale of compute have attracted increasi...
Relational graph neural networks have garnered particular attention to e...
Natural image matting is a fundamental and challenging computer vision t...
Real-world applications have high demands for semantic segmentation meth...
Pre-trained language models have achieved state-of-the-art results in va...
Distributed training has become a pervasive and effective approach for
t...
Deep neural networks (DNNs) exploit many layers and a large number of
pa...
The better accuracy and efficiency trade-off has been a challenging prob...
In recent years, image recognition applications have developed rapidly. ...
The task of image captioning aims to generate captions directly from ima...
Optical Character Recognition (OCR) systems have been widely used in var...
Pre-trained models have achieved state-of-the-art results in various Nat...
Being effective and efficient is essential to an object detector for
pra...
Recently, research efforts have been concentrated on revealing how
pre-t...
This paper studies compressing pre-trained language models, like BERT (D...
Automatic dialogue evaluation plays a crucial role in open-domain dialog...
Deep neural networks have achieved significant improvements in informati...