TextFuseNet: Scene Text Detection with Richer Fused Features

17 May 2020  ·  Jian Ye, Zhe Chen, Juhua Liu, Bo Du ·

Arbitrary shape text detection in natural scenes is an extremely challenging task. Unlike existing text detection approaches that only perceive texts based on limited feature representations, we propose a novel framework, namely TextFuseNet, to exploit the use of richer features fused for text detection. More specifically, we propose to perceive texts from three levels of feature representations, i.e., character-, word- and global-level, and then introduce a novel text representation fusion technique to help achieve robust arbitrary text detection. The multi-level feature representation can adequately describe texts by dissecting them into individual characters while still maintaining their general semantics. TextFuseNet then collects and merges the texts’ features from different levels using a multi-path fusion architecture which can effectively align and fuse different representations. In practice, our proposed TextFuseNet can learn a more adequate description of arbitrary shapes texts, suppressing false positives and producing more accurate detection results. Our proposed framework can also be trained with weak supervision for those datasets that lack character-level annotations. Experiments on several datasets show that the proposed TextFuseNet achieves state-of-the-art performance. Specifically, we achieve an F-measure of 94.3% on ICDAR2013, 92.1% on ICDAR2015, 87.1% on Total-Text and 86.6% on CTW-1500, respectively.

PDF Abstract

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Benchmark
Scene Text Detection IC19-Art TextFuseNet (ResNeXt-101) H-Mean 78.6 # 1
Scene Text Detection ICDAR 2013 TextFuseNet (ResNeXt-101) F-Measure 94.61% # 1
Precision 97.27 # 2
Recall 92.09 # 2
Scene Text Detection ICDAR 2015 TextFuseNet (ResNeXt-101) F-Measure 92.23 # 1
Precision 93.96 # 1
Recall 90.56 # 2
Scene Text Detection SCUT-CTW1500 TextFuseNet (ResNeXt-101) F-Measure 87.4 # 2
Precision 89.7 # 1
Recall 85.1 # 2
Scene Text Detection Total-Text TextFuseNet (ResNeXt-101) F-Measure 87.5% # 1
Precision 89.2 # 8
Recall 85.8 # 1

Methods


No methods listed for this paper. Add relevant methods here