---
task_categories:
- image-segmentation
- object-detection
- image-to-text
---
# gRefCOCO - Dataset for Generalized Referring Expression Segmentation, Comprehension, and Generation (GREx)
[](https://pytorch.org/)
[](https://www.python.org/downloads/)
[](https://paperswithcode.com/sota/generalized-referring-expression-segmentation?p=gres-generalized-referring-expression-1)
**[🏠[Project page]](https://henghuiding.github.io/GRES/)** **[📄[Paper (GREx)]](https://huggingface.co/papers/2601.05244)** **[📄[GRES Arxiv]](https://arxiv.org/abs/2306.00968)** **[📄[GREC Arxiv]](https://arxiv.org/abs/2308.16182)**
This repository contains information and tools for the **gRefCOCO** dataset, proposed in the paper:
> [GREx: Generalized Referring Expression Segmentation, Comprehension, and Generation](https://huggingface.co/papers/2601.05244)
> Chang Liu, Henghui Ding, Xudong Jiang
gRefCOCO is the first large-scale GREx dataset that contains multi-target, no-target, and single-target expressions and their corresponding images with labeled targets. It is designed to be backward-compatible with classic Referring Expression (REx) tasks.
## Usage
- Like RefCOCO, gRefCOCO also should be used together with images from the `train2014` of [MS COCO](https://cocodataset.org/#download).
- An example of dataloader [grefer.py](https://github.com/henghuiding/gRefCOCO/blob/main/grefer.py) is provided.
- We will update this repository with full API package and documentation soon. Please follow the usage in the [baseline code](https://github.com/henghuiding/ReLA) for now.
## Task 1 - GREC: Generalized Referring Expression Comprehension
- The GREC evaluation metric code is [here](https://github.com/henghuiding/gRefCOCO/blob/main/mdetr/datasets/refexp.py).
- We provide code based on [MDETR](https://github.com/ashkamath/mdetr), its training and inference are as follows:
### Training (Finetuning)
1. Process grefcoco to coco format.
```
python scripts/fine-tuning/grefexp_coco_format.py --data_path xxx --out_path mdetr_annotations/ --coco_path xxx
```
2. Training and download `pretrained_resnet101_checkpoint.pth` from [MDETR](https://github.com/ashkamath/mdetr)
```
python -m torch.distributed.launch --nproc_per_node=2 --use_env main.py --dataset_config configs/grefcoco.json --batch_size 4 --load pretrained_resnet101_checkpoint.pth --ema --text_encoder_lr 1e-5 --lr 5e-5 --output-dir grefcoco
```
### Inference
1. Obtain `checkpoint.pth` after training or download trained model [ here ☁️ Google Drive](https://drive.google.com/file/d/14OrM3n_Oap7xCT6nxj9QEnkJOUMpBGjB/view?usp=drive_link)
2. For test results, pass --test and --test_type test or testA or testB according to the dataset.
```
python -m torch.distributed.launch --nproc_per_node=2 --use_env main.py --dataset_config configs/grefcoco.json --batch_size 4 --resume grefcoco/checkpoint.pth --ema --eval
```
## Task 2 - GRES: Generalized Referring Expression Segmentation
Please refer to [ReLA](https://github.com/henghuiding/ReLA) for more details.
## Acknowledgement
Our project is built upon [refer](https://github.com/lichengunc/refer) and [cocoapi](https://github.com/cocodataset/cocoapi). Many thanks to the authors for their great works!
## BibTeX
Please consider to cite GRES/GREC/GREx if it helps your research.
```bibtex
@article{GREx,
title={{GREx}: Generalized Referring Expression Segmentation, Comprehension, and Generation},
author={Liu, Chang and Ding, Henghui and Jiang, Xudong},
journal={arXiv preprint arXiv:2601.05244},
year={2026}
}
@inproceedings{GRES,
title={{GRES}: Generalized Referring Expression Segmentation},
author={Liu, Chang and Ding, Henghui and Jiang, Xudong},
booktitle={CVPR},
year={2023}
}
@article{GREC,
title={{GREC}: Generalized Referring Expression Comprehension},
author={He, Shuting and Ding, Henghui and Liu, Chang and Jiang, Xudong},
journal={arXiv preprint arXiv:2308.16182},
year={2023}
}
```