Jun Gao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Jun Gao has a total of 11 co-authors. Here are the top co-authors by number of shared publications:
- Ziqiang Cao
Soochow University
3 shared publications - Yu Xie
Bilibili Inc. China
2 shared publications - Chuanliu Fan
Soochow University
1 shared publication - Hujiang Huang
Soochow University
1 shared publication - Jiaqing Fan
School of Computer Science and Technology, Soochow University, Suzhou, Jiangsu, China
1 shared publication - Qian Qiao
Bilibili Inc. China
1 shared publication
Latest Publications
- 28 December 2024
- Published by Springer Nature
- p. 425-438
Select and Order: Optimizing Few-Shot Image Classification with In-Context Learning
- 28 December 2024
- Published by Springer Nature
- p. 425-438
Select and Order: Optimizing Few-Shot Image Classification with In-Context Learning
- Open Access
- 28 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 10134-10143
DNTextSpotter: Arbitrary-Shaped Scene Text Spotting via Improved Denoising Training
More and more end-to-end text spotting methods based on Transformer architecture have demonstrated superior performance. These methods utilize a bipartite graph matching algorithm to perform one-to-one optimal matching between predicted objects and actual objects. However, the instability of bipartite graph matching can lead to inconsistent optimization targets, thereby affecting the training performance of the model. Existing literature applies denoising training to solve the problem of bipartite graph matching instability in object detection tasks. Unfortunately, this denoising training method cannot be directly applied to text spotting tasks, as these tasks need to perform irregular shape detection tasks and more complex text recognition tasks than classification. To address this issue, we propose a novel denoising training method (DNTextSpotter) for arbitrary-shaped text spotting. Specifically, we decompose the queries of the denoising part into noised positional and noised content queries. We use the four Bezier control points of the Bezier center curve to generate the noised positional queries. For the noised content queries, considering that the output of the text in a fixed positional order is not conducive to aligning position with content, we employ a masked character sliding method to initialize noised content queries, thereby assisting in the alignment of text content and position. Additionally, to improve the model's perception of the background, we further utilize an additional loss function for background characters classification in the denoising training part. DNTextSpotter outperforms state-of-the-art methods on four benchmarks-Total-Text, SCUT-CTW1500, ICDAR15, and Inverse-Text-most notably achieving an 11.3% improvement over the best approach on Inverse-Text.
- 28 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 10134-10143
Open AccessDNTextSpotter: Arbitrary-Shaped Scene Text Spotting via Improved Denoising Training
More and more end-to-end text spotting methods based on Transformer architecture have demonstrated superior performance. These methods utilize a bipartite graph matching algorithm to perform one-to-one optimal matching between predicted objects and actual objects. However, the instability of bipartite graph matching can lead to inconsistent optimization targets, thereby affecting the training performance of the model. Existing literature applies denoising training to solve the problem of bipartite graph matching instability in object detection tasks. Unfortunately, this denoising training method cannot be directly applied to text spotting tasks, as these tasks need to perform irregular shape detection tasks and more complex text recognition tasks than classification. To address this issue, we propose a novel denoising training method (DNTextSpotter) for arbitrary-shaped text spotting. Specifically, we decompose the queries of the denoising part into noised positional and noised content queries. We use the four Bezier control points of the Bezier center curve to generate the noised positional queries. For the noised content queries, considering that the output of the text in a fixed positional order is not conducive to aligning position with content, we employ a masked character sliding method to initialize noised content queries, thereby assisting in the alignment of text content and position. Additionally, to improve the model's perception of the background, we further utilize an additional loss function for background characters classification in the denoising training part. DNTextSpotter outperforms state-of-the-art methods on four benchmarks-Total-Text, SCUT-CTW1500, ICDAR15, and Inverse-Text-most notably achieving an 11.3% improvement over the best approach on Inverse-Text.
- 30 August 2024
- Published by Elsevier in Information Processing & Management
- Vol. 61 (6)
SelfCP: Compressing over-limit prompt via the frozen large language model itself
- 30 August 2024
- Published by Elsevier in Information Processing & Management
- Vol. 61 (6)
SelfCP: Compressing over-limit prompt via the frozen large language model itself