6 papers
MAM: Masked Acoustic Modeling for End-to-End Speech-to-Text Translation
Junkun Chen, Mingbo Ma, Renjie Zheng +1
End-to-end Speech-to-text Translation (E2E-ST), which directly translates source language speech to target language text, is widely useful in practice, but traditional cascaded app…
DropAttention: A Regularization Method for Fully-Connected Self-Attention Networks
Lin Zehui, Pengfei Liu, Luyao Huang +3
Variants dropout methods have been designed for the fully-connected layer, convolutional layer and recurrent layer in neural networks, and shown to be effective to avoid overfittin…
VATEX: A Large-Scale, High-Quality Multilingual Dataset for Video-and-Language Research
Xin Wang, Jiawei Wu, Junkun Chen +3
We present a new large-scale multilingual video description dataset, VATEX, which contains over 41,250 videos and 825,000 captions in both English and Chinese. Among the captions,…
Exploring Shared Structures and Hierarchies for Multiple NLP Tasks
Junkun Chen, Kaiyu Chen, Xinchi Chen +2
Designing shared neural architecture plays an important role in multi-task learning. The challenge is that finding an optimal sharing scheme heavily relies on the expert knowledge…
Same Representation, Different Attentions: Shareable Sentence Representation Learning from Multiple Tasks
Renjie Zheng, Junkun Chen, Xipeng Qiu
Distributed representation plays an important role in deep learning based natural language processing. However, the representation of a sentence often varies in different tasks, wh…
Meta Multi-Task Learning for Sequence Modeling
Junkun Chen, Xipeng Qiu, Pengfei Liu +1
Semantic composition functions have been playing a pivotal role in neural representation learning of text sequences. In spite of their success, most existing models suffer from the…