Sign in
Optimal Boxes: Boosting End-to-End Scene Text Recognition by Adjusting Annotated Bounding Boxes via Reinforcement Learning.
Jingqun Tang
Wenming Qian
Luchuan Song
Xiena Dong
Lan Li
Xiang Bai
Published in:
ECCV (28) (2022)
Keyphrases
</>
end to end
bounding box
reinforcement learning
scene text recognition
learning algorithm
dynamic programming
congestion control
image data
object segmentation
object categories
image representation
probabilistic model
computational complexity
multiscale
object detectors
training data
feature selection