Efficient Scene Text Detection with Textual Attention Tower
arXiv:2002.03741
Abstract
Scene text detection has received attention for years and achieved an impressive performance across various benchmarks. In this work, we propose an efficient and accurate approach to detect multioriented text in scene images. The proposed feature fusion mechanism allows us to use a shallower network to reduce the computational complexity. A self-attention mechanism is adopted to suppress false positive detections. Experiments on public benchmarks including ICDAR 2013, ICDAR 2015 and MSRA-TD500 show that our proposed approach can achieve better or comparable performances with fewer parameters and less computational cost.
arXiv commentAccepted by ICASSP 2020
References in corpus (6)
- ADADELTA: An Adaptive Learning Rate Method
- Generalised Dice overlap as a deep learning loss function for highly unbalanced segmentations
- TextBoxes++: A Single-Shot Oriented Scene Text Detector
- PVANET: Deep but Lightweight Neural Networks for Real-time Object Detection
- IncepText: A New Inception-Text Module with Deformable PSROI Pooling for Multi-Oriented Scene Text Detection
- MegDet: A Large Mini-Batch Object Detector