A weakly supervised text spotter using only transcription annotations is proposed, but it largely reproduces an earlier method without benchmarking against it.
Icdar 2015 competition on robust reading
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2024 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Hear the Scene: Audio-Enhanced Text Spotting
A weakly supervised text spotter using only transcription annotations is proposed, but it largely reproduces an earlier method without benchmarking against it.