Masahiro Kaneko
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
Masahiro Kaneko has a total of 13 co-authors. Here are the top co-authors by number of shared publications:
- Mamoru Komachi
Hitotsubashi University, 2-1 Naka, Kunitachi, Tokyo 186-8601, Japan
3 shared publications - Tosho Hirasawa
OMRON SINIC X Corp
2 shared publications Aizhan Imankulova
Graduate School of System Design, Tokyo Metropolitan University, Hino, Tokyo, Japan
1 shared publicationHaruhiko Kondo
Surgical Oncology Division, National Cancer Center Hospital , Tokyo, Japan
1 shared publicationHaruhiko Nakayama
Surgical Oncology Division, National Cancer Center Hospital , Tokyo, Japan
1 shared publicationHisao Asamura
Surgical Oncology Division, National Cancer Center Hospital , Tokyo, Japan
1 shared publication
Latest Publications
- Open Access
- 22 June 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 10, 67653-67668
Pre-Trained Word Embedding and Language Model Improve Multimodal Machine Translation: A Case Study in Multi30K
Multimodal machine translation (MMT) is an attractive application of neural machine translation (NMT) that is commonly incorporated with image information. However, the MMT models proposed thus far have only comparable or slightly better performance than their text-only counterparts. One potential cause of this infeasibility is a lack of large-scale data. Most previous studies mitigate this limitation by employing large-scale textual parallel corpora, which are more accessible than multimodal parallel corpora, in various ways. However, these corpora are still available on only a limited scale in low-resource language pairs or domains. In this study, we leveraged monolingual (or multimodal monolingual) corpora, which are available at scale in most languages and domains, to improve MMT models. Our approach follows that of previous unimodal works that use monolingual corpora to train the word embedding or language model and incorporate them into NMT systems. While these methods demonstrated the advantage of using pre-trained representations, there is still room for MMT models to improve. To this end, our system employs debiasing procedures for the word embedding and multimodal extension of the language model (visual-language model, VLM) to make better use of the pre-trained knowledge in the MMT task. The results of evaluations conducted on the de facto MMT dataset for the English–German translation indicate that the improvement obtained using well-tailored word embedding and VLM is approximately +1.84 BLEU and +1.63 BLEU, respectively. The evaluation on multiple language pairs reveals their adoptability across the languages. Beyond the success of our system, we also conducted an extensive analysis on VLM manipulation and showed promising areas for developing better MMT models by exploiting VLM; some benefits brought by either modality are missing, and MMT with VLM generates less fluent translations. Our code is available at https://github.com/toshohirasawa/mmt-with-monolingual-data.
- 22 June 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 10, 67653-67668
Open AccessPre-Trained Word Embedding and Language Model Improve Multimodal Machine Translation: A Case Study in Multi30K
Multimodal machine translation (MMT) is an attractive application of neural machine translation (NMT) that is commonly incorporated with image information. However, the MMT models proposed thus far have only comparable or slightly better performance than their text-only counterparts. One potential cause of this infeasibility is a lack of large-scale data. Most previous studies mitigate this limitation by employing large-scale textual parallel corpora, which are more accessible than multimodal parallel corpora, in various ways. However, these corpora are still available on only a limited scale in low-resource language pairs or domains. In this study, we leveraged monolingual (or multimodal monolingual) corpora, which are available at scale in most languages and domains, to improve MMT models. Our approach follows that of previous unimodal works that use monolingual corpora to train the word embedding or language model and incorporate them into NMT systems. While these methods demonstrated the advantage of using pre-trained representations, there is still room for MMT models to improve. To this end, our system employs debiasing procedures for the word embedding and multimodal extension of the language model (visual-language model, VLM) to make better use of the pre-trained knowledge in the MMT task. The results of evaluations conducted on the de facto MMT dataset for the English–German translation indicate that the improvement obtained using well-tailored word embedding and VLM is approximately +1.84 BLEU and +1.63 BLEU, respectively. The evaluation on multiple language pairs reveals their adoptability across the languages. Beyond the success of our system, we also conducted an extensive analysis on VLM manipulation and showed promising areas for developing better MMT models by exploiting VLM; some benefits brought by either modality are missing, and MMT with VLM generates less fluent translations. Our code is available at https://github.com/toshohirasawa/mmt-with-monolingual-data.
- Open Access
- 1 January 2021
- Published by Association for Computational Linguistics (ACL)
- p. 143-149
Sentence Concatenation Approach to Data Augmentation for Neural Machine Translation
Seiichiro Kondo, Kengo Hotate, Tosho Hirasawa, Masahiro Kaneko, Mamoru Komachi. Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Student Research Workshop. 2021.
- 1 January 2021
- Published by Association for Computational Linguistics (ACL)
- p. 143-149
Open AccessSentence Concatenation Approach to Data Augmentation for Neural Machine Translation
Seiichiro Kondo, Kengo Hotate, Tosho Hirasawa, Masahiro Kaneko, Mamoru Komachi. Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Student Research Workshop. 2021.
- Open Access
- 15 September 2020
- Published by Association for Natural Language Processing in Journal of Natural Language Processing
- Vol. 27 (3) , 683-687
Encoder-Decoder Models Can Benefit from Pre-trained Masked Language Models in Grammatical Error Correction
- 15 September 2020
- Published by Association for Natural Language Processing in Journal of Natural Language Processing
- Vol. 27 (3) , 683-687
Open AccessEncoder-Decoder Models Can Benefit from Pre-trained Masked Language Models in Grammatical Error Correction
- Open Access
- 1 January 2018
- Published by Association for Computational Linguistics (ACL)
- p. 365-369
TMU System for SLAM-2018
- 1 January 2018
- Published by Association for Computational Linguistics (ACL)
- p. 365-369
Open AccessTMU System for SLAM-2018
- Open Access
- 1 June 1997
- Published by Oxford University Press (OUP) in Japanese Journal of Clinical Oncology
- Vol. 27 (3) , 204-205
CT-guided bronchoscopic barium marking for resection of a fluoroscopically invisible peripheral pulmonary lesion
Fluoroscopically invisible peripheral pulmonary lesions are increasingly detected by CT. Thoracoscopic resection is one option for histo-cytologic diagnosis of these lesions. Its diagnostic accuracy approximates to 100%. However, localization of such lesions is difficult thoracoscopically. Present localization aids include transthoracic needle implantation of wires. However, this carries the risk of pneumothorax and wire dislodgment. We therefore performed transbronchial placement of a barium sulfate suspension to serve as a fluoroscopic marker for a fluoroscopically invisible peripheral pulmonary lesion. This was done via CT-guided bronchoscopy (1).
- 1 June 1997
- Published by Oxford University Press (OUP) in Japanese Journal of Clinical Oncology
- Vol. 27 (3) , 204-205
Open AccessCT-guided bronchoscopic barium marking for resection of a fluoroscopically invisible peripheral pulmonary lesion
Fluoroscopically invisible peripheral pulmonary lesions are increasingly detected by CT. Thoracoscopic resection is one option for histo-cytologic diagnosis of these lesions. Its diagnostic accuracy approximates to 100%. However, localization of such lesions is difficult thoracoscopically. Present localization aids include transthoracic needle implantation of wires. However, this carries the risk of pneumothorax and wire dislodgment. We therefore performed transbronchial placement of a barium sulfate suspension to serve as a fluoroscopic marker for a fluoroscopically invisible peripheral pulmonary lesion. This was done via CT-guided bronchoscopy (1).