The heavily occluded scene text (HOST) dataset is a dataset that contains images of text with occlusions. It is used to improve the recognition performance of occluded text in machine vision applications 1. The dataset is composed of 4832 images that are manually occluded in weak or heavy degrees.
Variants: HOST
This dataset is used in 1 benchmark:
| Task | Model | Paper | Date |
|---|---|---|---|
| Scene Text Recognition | CLIP4STR-L | CLIP4STR: A Simple Baseline for … | 2023-05-23 |
| Scene Text Recognition | CLIP4STR-B | CLIP4STR: A Simple Baseline for … | 2023-05-23 |
| Scene Text Recognition | CCD-ViT-Base | Self-supervised Character-to-Character Distillation for Text … | 2022-11-01 |
Recent papers with results on this dataset: