Title | ||
---|---|---|
Enhancing Neural Sign Language Translation by highlighting the facial expression information |
Abstract | ||
---|---|---|
•To the best of our knowledge, this is the first work that explicitly explores facial expressions in neural SLT task.•Non-independent Multi-stream Architecture as SFoI-Net-FHM is introduced, which can independently extract facial and global features to consistently fine-grained features.•A RoIs-based Multi-region Architecture as alternative SFoI-Net-FHM is proposed. Under low resource conditions, this is the first work of converting RoIs exploration of SLT into object detection tasks. |
Year | DOI | Venue |
---|---|---|
2021 | 10.1016/j.neucom.2021.08.079 | Neurocomputing |
Keywords | DocType | Volume |
Sign Language Translation,Facial Expression Information,Non-independent Multi-stream Architecture,RoIs-based Multi-region Architecture | Journal | 464 |
ISSN | Citations | PageRank |
0925-2312 | 0 | 0.34 |
References | Authors | |
0 | 5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Jiangbin Zheng | 1 | 0 | 0.68 |
CHEN Yi-dong | 2 | 106 | 27.34 |
Chong Wu | 3 | 55 | 9.61 |
Xiaodong Shi | 4 | 8 | 2.16 |
Suhail Muhammad Kamal | 5 | 0 | 0.34 |