The lip reading method based on Adaptive Pooling Attention Transformer
Lip reading technology establishes the mapping relationship between lip movements and specific language characters by processing a series of consecutive lip images, thereby enabling semantic information recognition. Existing methods mainly use recurrent networks for spatiotemporal modeling of sequen...
Saved in:
| Main Authors: | YAO Yun, HU Zhenxiao, DENG Tao, WANG Xiao |
|---|---|
| Format: | Article |
| Language: | zho |
| Published: |
POSTS&TELECOM PRESS Co., LTD
2025-01-01
|
| Series: | 智能科学与技术学报 |
| Subjects: | |
| Online Access: | http://www.cjist.com.cn/zh/article/99639204/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Similar Items
-
A lip reading method based on adaptive pooling attention Transformer
by: YAO Yun, et al.
Published: (2025-06-01) -
Molten Pool Image Segmentation Based on Adaptive Multi-Scale Attention Mechanism
by: Yuefeng Chen, et al.
Published: (2025-01-01) -
Graph-based vision transformer with sparsity for training on small datasets from scratch
by: Peng Li, et al.
Published: (2025-07-01) -
SAFH-Net: A Hybrid Network With Shuffle Attention and Adaptive Feature Fusion for Enhanced Retinal Vessel Segmentation
by: Yang Zhou Ling Ou, et al.
Published: (2025-01-01) -
Invertible Attention-Guided Adaptive Convolution and Dual-Domain Transformer for Pansharpening
by: Qun Song, et al.
Published: (2025-01-01)