Huizhen Wang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Huizhen Wang has a total of 16 co-authors. Here are the top co-authors by number of shared publications:
- Jingbo Zhu
School of Computer Science and Engineering Northeastern University Shenyang China
6 shared publications - Muhua Zhu
Alibaba Group
4 shared publications Tong Xiao
Northeastern University
2 shared publications- Baoyu Hou
Northeastern University
1 shared publication Bei Li
School of Computer Science and Engineering NLP Lab Northeastern University Shenyang China
1 shared publicationFeiliang Ren
1 shared publication
Latest Publications
- 1 November 2024
- Published by Springer Nature
- p. 420-432
DFS-QA: Dynamic Frame Selection for Better Video Question Answering
- 1 November 2024
- Published by Springer Nature
- p. 420-432
DFS-QA: Dynamic Frame Selection for Better Video Question Answering
- 20 September 2023
- Published by Springer Nature
- p. 58-71
Overcoming Language Priors with Counterfactual Inference for Visual Question Answering
- 20 September 2023
- Published by Springer Nature
- p. 58-71
Overcoming Language Priors with Counterfactual Inference for Visual Question Answering
- Open Access
Shallow-to-Deep Training for Neural Machine Translation
Deep encoders have been proven to be effective in improving neural machine translation (NMT) systems, but training an extremely deep encoder is time consuming. Moreover, why deep models help NMT is an open question. In this paper, we investigate the behavior of a well-tuned deep Transformer system. We find that stacking layers is helpful in improving the representation ability of NMT models and adjacent layers perform similarly. This inspires us to develop a shallow-to-deep training method that learns deep models by stacking shallow models. In this way, we successfully train a Transformer system with a 54-layer encoder. Experimental results on WMT'16 English-German and WMT'14 English-French translation tasks show that it is $1.4$ $\times$ faster than training from scratch, and achieves a BLEU score of $30.33$ and $43.29$ on two tasks. The code is publicly available at https://github.com/libeineu/SDT-Training/.Open AccessShallow-to-Deep Training for Neural Machine Translation
Deep encoders have been proven to be effective in improving neural machine translation (NMT) systems, but training an extremely deep encoder is time consuming. Moreover, why deep models help NMT is an open question. In this paper, we investigate the behavior of a well-tuned deep Transformer system. We find that stacking layers is helpful in improving the representation ability of NMT models and adjacent layers perform similarly. This inspires us to develop a shallow-to-deep training method that learns deep models by stacking shallow models. In this way, we successfully train a Transformer system with a 54-layer encoder. Experimental results on WMT'16 English-German and WMT'14 English-French translation tasks show that it is $1.4$ $\times$ faster than training from scratch, and achieves a BLEU score of $30.33$ and $43.29$ on two tasks. The code is publicly available at https://github.com/libeineu/SDT-Training/. - 20 July 2013
- Published by China Science Publishing & Media Ltd. in Journal of Software
- Vol. 24 (7) , 1545-1556
Optimizations of Multi-Aspect Rating Inference Model
- 20 July 2013
- Published by China Science Publishing & Media Ltd. in Journal of Software
- Vol. 24 (7) , 1545-1556
Optimizations of Multi-Aspect Rating Inference Model
- 19 June 2013
- Published by Cambridge University Press (CUP) in Natural Language Engineering
- Vol. 21 (1) , 113-138
Improving shift-reduce constituency parsing with large-scale unlabeled data
Shift-reduce parsing has been studied extensively for diverse grammars due to the simplicity and running efficiency. However, in the field of constituency parsing, shift-reduce parsers lag behind state-of-the-art parsers. In this paper we propose a semi-supervised approach for advancing shift-reduce constituency parsing. First, we apply the uptraining approach (Petrov, S. et al. 2010. In Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing (EMNLP), Cambridge, MA, USA, pp. 705–713) to improve part-of-speech taggers to provide better part-of-speech tags to subsequent shift-reduce parsers. Second, we enhance shift-reduce parsing models with novel features that are defined on lexical dependency information. Both stages depend on the use of large-scale unlabeled data. Experimental results show that the approach achieves overall improvements of 1.5 percent and 2.1 percent on English and Chinese data respectively. Moreover, the final parsing accuracies reach 90.9 percent and 82.2 percent respectively, which are comparable with the accuracy of state-of-the-art parsers.
- 19 June 2013
- Published by Cambridge University Press (CUP) in Natural Language Engineering
- Vol. 21 (1) , 113-138
Improving shift-reduce constituency parsing with large-scale unlabeled data
Shift-reduce parsing has been studied extensively for diverse grammars due to the simplicity and running efficiency. However, in the field of constituency parsing, shift-reduce parsers lag behind state-of-the-art parsers. In this paper we propose a semi-supervised approach for advancing shift-reduce constituency parsing. First, we apply the uptraining approach (Petrov, S. et al. 2010. In Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing (EMNLP), Cambridge, MA, USA, pp. 705–713) to improve part-of-speech taggers to provide better part-of-speech tags to subsequent shift-reduce parsers. Second, we enhance shift-reduce parsing models with novel features that are defined on lexical dependency information. Both stages depend on the use of large-scale unlabeled data. Experimental results show that the approach achieves overall improvements of 1.5 percent and 2.1 percent on English and Chinese data respectively. Moreover, the final parsing accuracies reach 90.9 percent and 82.2 percent respectively, which are comparable with the accuracy of state-of-the-art parsers.
- 31 May 2013
- Published by Springer Nature in Language Resources and Evaluation
- Vol. 47 (4) , 1213-1231
A feature-based approach to better automatic treebank conversion
- 31 May 2013
- Published by Springer Nature in Language Resources and Evaluation
- Vol. 47 (4) , 1213-1231
A feature-based approach to better automatic treebank conversion