---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: segment_mask
    dtype: image
  - name: detections
    dtype: string
  - name: referring_expression
    dtype: string
  splits:
  - name: train
    num_bytes: 17651696.0
    num_examples: 100
  download_size: 17392053
  dataset_size: 17651696.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---

### Paper Link

[**SOS: Synthetic Object Segments Improve Detection, Segmentation, and Grounding**](https://arxiv.org/abs/2510.09110)

### Code repo 

[**Code for Generation**](https://github.com/weikaih04/Synthetic-Detection-Segmentation-Grounding-Data)


### Citation

```bibtex
@misc{huang2025sossyntheticobjectsegments,
  title={SOS: Synthetic Object Segments Improve Detection, Segmentation, and Grounding},
  author={Weikai Huang and Jieyu Zhang and Taoyang Jia and Chenhao Zheng and Ziqi Gao and Jae Sung Park and Ranjay Krishna},
  year={2025},
  eprint={2510.09110},
  archivePrefix={arXiv},
  primaryClass={cs.CV},
  url={https://arxiv.org/abs/2510.09110}
}