Masato Fujitake
This paper presents Diffusion Model for Scene Text Recognition (DiffusionSTR), an end-to-end text recognition framework using diffusion models for recognizing text in the wild. While existing studies have viewed the scene text recognition task as an image-to-text transformation, we rethought it as a text-text one under images in a diffusion model. We show for the first time that the diffusion model can be applied to text recognition. Furthermore, experimental results on publicly available datasets show that the proposed method achieves competitive accuracy compared to state-of-the-art methods.
| Task | Dataset | Metric | Value | Model |
|---|---|---|---|---|
| Scene Parsing | SVT | Accuracy | 93.6 | DiffusionSTR |
| Scene Parsing | SVTP | Accuracy | 89.2 | DiffusionSTR |
| Scene Parsing | CUTE80 | Accuracy | 92.5 | DiffusionSTR |
| Scene Parsing | ICDAR2015 | Accuracy | 86 | DiffusionSTR |
| Scene Parsing | IIIT5k | Accuracy | 97.3 | DiffusionSTR |
| Scene Parsing | ICDAR2013 | Accuracy | 97.1 | DiffusionSTR |
| 2D Semantic Segmentation | SVT | Accuracy | 93.6 | DiffusionSTR |
| 2D Semantic Segmentation | SVTP | Accuracy | 89.2 | DiffusionSTR |
| 2D Semantic Segmentation | CUTE80 | Accuracy | 92.5 | DiffusionSTR |
| 2D Semantic Segmentation | ICDAR2015 | Accuracy | 86 | DiffusionSTR |
| 2D Semantic Segmentation | IIIT5k | Accuracy | 97.3 | DiffusionSTR |
| 2D Semantic Segmentation | ICDAR2013 | Accuracy | 97.1 | DiffusionSTR |
| Scene Text Recognition | SVT | Accuracy | 93.6 | DiffusionSTR |
| Scene Text Recognition | SVTP | Accuracy | 89.2 | DiffusionSTR |
| Scene Text Recognition | CUTE80 | Accuracy | 92.5 | DiffusionSTR |
| Scene Text Recognition | ICDAR2015 | Accuracy | 86 | DiffusionSTR |
| Scene Text Recognition | IIIT5k | Accuracy | 97.3 | DiffusionSTR |
| Scene Text Recognition | ICDAR2013 | Accuracy | 97.1 | DiffusionSTR |