Test results in MSRA-TD500 dataset.

<div><p>With the continuous advancement of deep learning, research in scene text detection has evolved significantly. However, complex backgrounds and various text forms complicate the task of detecting text from images. CNN is a deep learning algorithm that automatically extracts featur...

Full description

Saved in:

Bibliographic Details
Main Author:	Yuan Li (67017) (author)
Published:	2024
Subjects:	Microbiology Science Policy Space Science Biological Sciences not elsewhere classified Mathematical Sciences not elsewhere classified Information Systems not elsewhere classified spatial enhanced self integration effectively facilitates channel enhanced self automatically extracts features achieves performance improvements lacks global attributes feature decoder designed icdar 2015 dataset global contextual information dual perspective cnn scene text detection paper significantly improves deep learning algorithm transformer </ p global information text detection feature map evolved significantly deep learning text dataset td500 dataset detailed information xlink "> thereby alleviating recent years positional relationships paper introduces experiments show different types detecting text convolution operation continuous advancement computer vision complex backgrounds attention module 72 vs 07 vs
Tags:	Add Tag No Tags, Be the first to tag this record!

Test results in MSRA-TD500 dataset.

Similar Items