---
license: cc-by-4.0
task_categories:
- image-feature-extraction
---

Licenses for existing assets:

This dataset is used for the paper [TokBench: Evaluating Your Visual Tokenizer before Visual Generation](https://huggingface.co/papers/2505.18142).
Project page: https://wjf5203.github.io/TokBench

Total-Text: BSD-3-CLause

TextOCR: CC BY 4.0

SROIE：MIT

CORD： CC BY 4.0

DocVQA： MIT

ICDAR 2013： MIT

ICDAR 2015： MIT

Infographic: MIT

WFLW: Apache 2.0