Get our free extension to see links to code for papers anywhere online!

Add to Chrome

Add to Firefox

Picture for Taro Watanabe

Taro Watanabe

Switching to Discriminative Image Captioning by Relieving a Bottleneck of Reinforcement Learning

Dec 06, 2022
Ukyo Honda, Taro Watanabe, Yuji Matsumoto

* WACV 2023 (19 pages, 9 figures)

Access Paper or Ask Questions

$N$-gram Is Back: Residual Learning of Neural Text Generation with $n$-gram Language Model

Nov 03, 2022
Huayang Li, Deng Cai, Jin Xu, Taro Watanabe

Figure 1 for $N$-gram Is Back: Residual Learning of Neural Text Generation with $n$-gram Language Model

Figure 2 for $N$-gram Is Back: Residual Learning of Neural Text Generation with $n$-gram Language Model

Figure 3 for $N$-gram Is Back: Residual Learning of Neural Text Generation with $n$-gram Language Model

Figure 4 for $N$-gram Is Back: Residual Learning of Neural Text Generation with $n$-gram Language Model

* Accepted to findings of EMNLP 2022

Access Paper or Ask Questions

Adapting to Non-Centered Languages for Zero-shot Multilingual Translation

Sep 09, 2022
Zhi Qu, Taro Watanabe

Figure 1 for Adapting to Non-Centered Languages for Zero-shot Multilingual Translation

Figure 2 for Adapting to Non-Centered Languages for Zero-shot Multilingual Translation

Figure 3 for Adapting to Non-Centered Languages for Zero-shot Multilingual Translation

Figure 4 for Adapting to Non-Centered Languages for Zero-shot Multilingual Translation

* Accepted at COLING 2022

Access Paper or Ask Questions

Visualizing the Relationship Between Encoded Linguistic Information and Task Performance

Mar 29, 2022
Jiannan Xiang, Huayang Li, Defu Lian, Guoping Huang, Taro Watanabe, Lemao Liu

Figure 1 for Visualizing the Relationship Between Encoded Linguistic Information and Task Performance

Figure 2 for Visualizing the Relationship Between Encoded Linguistic Information and Task Performance

Figure 3 for Visualizing the Relationship Between Encoded Linguistic Information and Task Performance

Figure 4 for Visualizing the Relationship Between Encoded Linguistic Information and Task Performance

* Findings of ACL 2022

Access Paper or Ask Questions

Transductive Data Augmentation with Relational Path Rule Mining for Knowledge Graph Embedding

Nov 01, 2021
Yushi Hirose, Masashi Shimbo, Taro Watanabe

Figure 1 for Transductive Data Augmentation with Relational Path Rule Mining for Knowledge Graph Embedding

Figure 2 for Transductive Data Augmentation with Relational Path Rule Mining for Knowledge Graph Embedding

Figure 3 for Transductive Data Augmentation with Relational Path Rule Mining for Knowledge Graph Embedding

Figure 4 for Transductive Data Augmentation with Relational Path Rule Mining for Knowledge Graph Embedding

* 8 pages, 0 figures, accepted by 2021 IEEE International Conference on Big Knowledge. The copyright of this paper has been transferred to the IEEE, please comply with the copyright of the IEEE

Access Paper or Ask Questions

Removing Word-Level Spurious Alignment between Images and Pseudo-Captions in Unsupervised Image Captioning

Apr 28, 2021
Ukyo Honda, Yoshitaka Ushiku, Atsushi Hashimoto, Taro Watanabe, Yuji Matsumoto

Figure 1 for Removing Word-Level Spurious Alignment between Images and Pseudo-Captions in Unsupervised Image Captioning

Figure 2 for Removing Word-Level Spurious Alignment between Images and Pseudo-Captions in Unsupervised Image Captioning

Figure 3 for Removing Word-Level Spurious Alignment between Images and Pseudo-Captions in Unsupervised Image Captioning

Figure 4 for Removing Word-Level Spurious Alignment between Images and Pseudo-Captions in Unsupervised Image Captioning

* EACL 2021 (11 pages, 3 figures; corrected the description of pseudo-caption preprocessing)

Access Paper or Ask Questions

Denoising Neural Machine Translation Training with Trusted Data and Online Data Selection

Aug 31, 2018
Wei Wang, Taro Watanabe, Macduff Hughes, Tetsuji Nakagawa, Ciprian Chelba

Figure 1 for Denoising Neural Machine Translation Training with Trusted Data and Online Data Selection

Figure 2 for Denoising Neural Machine Translation Training with Trusted Data and Online Data Selection

Figure 3 for Denoising Neural Machine Translation Training with Trusted Data and Online Data Selection

Figure 4 for Denoising Neural Machine Translation Training with Trusted Data and Online Data Selection

* 11 pages, 2018 Third Conference on Machine Translation (WMT18)

Access Paper or Ask Questions