A Fast Scene Text Detector Using Knowledge Distillation
Incidental scene text detection is a challenging problem because of arbitrary orientation, low resolution, perspective distortion, and variant aspect ratios of text in natural images. In this paper, we present an end-to-end trainable deep model, which can effectively and efficiently locate multi-ori...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2019-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/8626192/ |
| _version_ | 1828967018202660864 |
|---|---|
| author | Peng Yang Fanlong Zhang Guowei Yang |
| author_facet | Peng Yang Fanlong Zhang Guowei Yang |
| author_sort | Peng Yang |
| collection | DOAJ |
| description | Incidental scene text detection is a challenging problem because of arbitrary orientation, low resolution, perspective distortion, and variant aspect ratios of text in natural images. In this paper, we present an end-to-end trainable deep model, which can effectively and efficiently locate multi-oriented scene text. Our detector includes a student network and a teacher network, and they inherit complex VGGNet and lightweight PVANet architecture, respectively. While deploying for text detection, the teacher network is used to guide the training process of a student via knowledge distilling so as to maintain the tradeoff between accuracy and efficiency. We have evaluated the proposed detector on three popular benchmarks, and it achieves F-measures of 83.7%, 57.27%, and 90% on ICDAR2015 Incidental Scene Text, COCO-Text, and ICDAR2013, respectively, which outperforms the most state-of-the-art methods. |
| first_indexed | 2024-12-14T11:42:59Z |
| format | Article |
| id | doaj.art-9c5c5fc306684054b245ab16cf0b9e9d |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-12-14T11:42:59Z |
| publishDate | 2019-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-9c5c5fc306684054b245ab16cf0b9e9d2022-12-21T23:02:44ZengIEEEIEEE Access2169-35362019-01-017225882259810.1109/ACCESS.2019.28953308626192A Fast Scene Text Detector Using Knowledge DistillationPeng Yang0https://orcid.org/0000-0002-1505-7857Fanlong Zhang1https://orcid.org/0000-0001-8865-9683Guowei Yang2School of Information Engineering, Nanjing Audit University, Jiangshu, ChinaSchool of Information Engineering, Nanjing Audit University, Jiangshu, ChinaSchool of Information Engineering, Nanjing Audit University, Jiangshu, ChinaIncidental scene text detection is a challenging problem because of arbitrary orientation, low resolution, perspective distortion, and variant aspect ratios of text in natural images. In this paper, we present an end-to-end trainable deep model, which can effectively and efficiently locate multi-oriented scene text. Our detector includes a student network and a teacher network, and they inherit complex VGGNet and lightweight PVANet architecture, respectively. While deploying for text detection, the teacher network is used to guide the training process of a student via knowledge distilling so as to maintain the tradeoff between accuracy and efficiency. We have evaluated the proposed detector on three popular benchmarks, and it achieves F-measures of 83.7%, 57.27%, and 90% on ICDAR2015 Incidental Scene Text, COCO-Text, and ICDAR2013, respectively, which outperforms the most state-of-the-art methods.https://ieeexplore.ieee.org/document/8626192/Scene text detectionmulti-oriented textdeep neural networkknowledge distilling |
| spellingShingle | Peng Yang Fanlong Zhang Guowei Yang A Fast Scene Text Detector Using Knowledge Distillation IEEE Access Scene text detection multi-oriented text deep neural network knowledge distilling |
| title | A Fast Scene Text Detector Using Knowledge Distillation |
| title_full | A Fast Scene Text Detector Using Knowledge Distillation |
| title_fullStr | A Fast Scene Text Detector Using Knowledge Distillation |
| title_full_unstemmed | A Fast Scene Text Detector Using Knowledge Distillation |
| title_short | A Fast Scene Text Detector Using Knowledge Distillation |
| title_sort | fast scene text detector using knowledge distillation |
| topic | Scene text detection multi-oriented text deep neural network knowledge distilling |
| url | https://ieeexplore.ieee.org/document/8626192/ |
| work_keys_str_mv | AT pengyang afastscenetextdetectorusingknowledgedistillation AT fanlongzhang afastscenetextdetectorusingknowledgedistillation AT guoweiyang afastscenetextdetectorusingknowledgedistillation AT pengyang fastscenetextdetectorusingknowledgedistillation AT fanlongzhang fastscenetextdetectorusingknowledgedistillation AT guoweiyang fastscenetextdetectorusingknowledgedistillation |