EAFormer: Scene Text Segmentation with Edge-Aware Transformers.

Haiyang Yu Teng Fu Bin Li Xiangyang Xue

Published in: CoRR (2024)

Keyphrases

text segmentation
video sequences
dynamic scenes
image sequences
edge detection
sentence level
multiple objects
edge information
document set
moving objects
single image
computer vision
character recognition
keywords
text mining
text classification
target object
graph model
complex scenes