537 citations · 902 across the 19 of their papers we have counts for
19 papers
TranSFormer: Slow-Fast Transformer for Machine Translation
Bei Li, Yi Jing, Xu Tan +3
Learning multiscale Transformer models has been evidenced as a viable approach to augmenting machine translation systems. Prior research has primarily focused on treating subwords…
Towards Understanding Omission in Dialogue Summarization
Yicheng Zou, Kaitao Song, Xu Tan +4
Dialogue summarization aims to condense the lengthy dialogue into a concise summary, and has recently achieved significant progress. However, the result of existing methods is stil…
InferGrad: Improving Diffusion Models for Vocoder by Considering Inference in Training
Zehua Chen, Xu Tan, Ke Wang +4
Denoising diffusion probabilistic models (diffusion models for short) require a large number of iterations in inference to achieve the generation quality that matches or surpasses…
AdaSpeech 3: Adaptive Text to Speech for Spontaneous Style
Yuzi Yan, Xu Tan, Bohan Li +6
While recent text to speech (TTS) models perform very well in synthesizing reading-style (e.g., audiobook) speech, it is still challenging to synthesize spontaneous-style speech (e…
Libri-adhoc40: A dataset collected from synchronized ad-hoc microphone arrays
Shanzheng Guan, Shupei Liu, Junqi Chen +8
Recently, there is a research trend on ad-hoc microphone arrays. However, most research was conducted on simulated data. Although some data sets were collected with a small number…
Cross-domain Speech Recognition with Unsupervised Character-level Distribution Matching
Wenxin Hou, Jindong Wang, Xu Tan +2
End-to-end automatic speech recognition (ASR) can achieve promising performance with large-scale training data. However, it is known that domain mismatch between training and testi…
Adaptive Logit Adjustment Loss for Long-Tailed Visual Recognition
Yan Zhao, Weicong Chen, Xu Tan +2
Data in the real world tends to exhibit a long-tailed label distribution, which poses great challenges for the training of neural networks in visual recognition. Existing methods t…
BRECQ: Pushing the Limit of Post-Training Quantization by Block Reconstruction
Yuhang Li, Ruihao Gong, Xu Tan +6
We study the challenging task of neural network quantization without end-to-end retraining, called Post-training Quantization (PTQ). PTQ usually requires a small subset of training…
Speech enhancement aided end-to-end multi-task learning for voice activity detection
Xu Tan, Xiao-Lei Zhang
Robust voice activity detection (VAD) is a challenging task in low signal-to-noise (SNR) environments. Recent studies show that speech enhancement is helpful to VAD, but the perfor…
LightPAFF: A Two-Stage Distillation Framework for Pre-training and Fine-tuning
Kaitao Song, Hao Sun, Xu Tan +4
While pre-training and fine-tuning, e.g., BERT~\citep{devlin2018bert}, GPT-2~\citep{radford2019language}, have achieved great success in language understanding and generation tasks…
MPNet: Masked and Permuted Pre-training for Language Understanding
Kaitao Song, Xu Tan, Tao Qin +2
BERT adopts masked language modeling (MLM) for pre-training and is one of the most successful pre-training models. Since BERT neglects dependency among predicted tokens, XLNet intr…
Representation Degeneration Problem in Training Natural Language Generation Models
Jun Gao, Di He, Xu Tan +3
We study an interesting problem in training neural network-based models for natural language generation tasks, which we call the \emph{representation degeneration problem}. We obse…
Almost Unsupervised Text to Speech and Automatic Speech Recognition
Yi Ren, Xu Tan, Tao Qin +3
Text to speech (TTS) and automatic speech recognition (ASR) are two dual tasks in speech processing and both achieve impressive performance thanks to the recent advance in deep lea…
Token-Level Ensemble Distillation for Grapheme-to-Phoneme Conversion
Hao Sun, Xu Tan, Jun-Wei Gan +4
Grapheme-to-phoneme (G2P) conversion is an important task in automatic speech recognition and text-to-speech systems. Recently, G2P conversion is viewed as a sequence to sequence t…
Non-Autoregressive Neural Machine Translation with Enhanced Decoder Input
Junliang Guo, Xu Tan, Di He +3
Non-autoregressive translation (NAT) models, which remove the dependence on previous target tokens from the inputs of the decoder, achieve significantly inference speedup but at th…
Sentence-wise Smooth Regularization for Sequence to Sequence Learning
Chengyue Gong, Xu Tan, Di He +1
Maximum-likelihood estimation (MLE) is widely used in sequence to sequence tasks for model training. It uniformly treats the generation/prediction of each target token as multi-cla…
FRAGE: Frequency-Agnostic Word Representation
Chengyue Gong, Di He, Xu Tan +3
Continuous word representation (aka word embedding) is a basic building block in many neural network-based models used in natural language processing tasks. Although it is widely a…
Double Path Networks for Sequence to Sequence Learning
Kaitao Song, Xu Tan, Di He +3
Encoder-decoder based Sequence to Sequence learning (S2S) has made remarkable progress in recent years. Different network architectures have been used in the encoder/decoder. Among…
Dense Information Flow for Neural Machine Translation
Yanyao Shen, Xu Tan, Di He +2
Recently, neural machine translation has achieved remarkable progress by introducing well-designed deep neural networks into its encoder-decoder framework. From the optimization pe…