WaterMask - GitLite

KO | EN

gitlite — search

#javascript #typescript #golang #android #deep-learning #java #windows #cli #kubernetes #awesome-list #pytorch #security #ios #open-source

WaterMask

(★ 80)

This repo is the official implementation of "WaterMask: Instance Segmentation for Underwater Imagery", and has been accepted by ICCV2023! 🎉🎉🎉

File Explorer

Download Latest Version (.zip)

- framework.png
- uiis_dataset.png
- - __init__.py
- - analyze_logs.py
  - analyze_results.py
  - benchmark.py
  - browse_dataset.py
  - browse_grounding_dataset.py
  - browse_grounding_raw.py
  - coco_error_analysis.py
  - coco_occluded_separated_recall.py
  - confusion_matrix.py
  - eval_metric.py
  - fuse_results.py
  - get_flops.py
  - optimize_anchors.py
  - robustness_eval.py
  - test_robustness.py
- - ade20k2coco.py
  - cityscapes.py
  - coco2odvg.py
  - coco2ovd.py
  - coco_stuff164k.py
  - crowdhuman2coco.py
  - extract_coco_from_mixed.py
  - fix_o365_names.py
  - goldg2odvg.py
  - grit2odvg.py
  - grit_processing.py
  - images2coco.py
  - lvis2odvg.py
  - lvis2ovd.py
  - mot2coco.py
  - mot2reid.py
  - openimages2odvg.py
  - pascal_voc.py
  - prepare_coco_semantic_annos_from_panoptic_annos.py
  - refcoco2odvg.py
  - remove_cocotrain2017_from_refcoco.py
  - youtubevis2coco.py
  - zhiyuan_objv2_train_names_fix.csv
- - mmdet2torchserve.py
  - mmdet_handler.py
  - test_torchserver.py
- - download_dataset.py
  - gen_coco_panoptic_test_info.py
  - get_crowdhuman_id_hw.py
  - get_image_metas.py
  - print_config.py
  - split_coco.py
  - split_odvg.py
- - detectron2_to_mmdet.py
  - detectron2pytorch.py
  - detic_to_mmdet.py
  - glip_to_mmdet.py
  - groundingdino_to_mmdet.py
  - publish_model.py
  - regnet2mmdet.py
  - selfsup2mmdet.py
  - swinv1_to_mmdet.py
  - upgrade_model_version.py
  - upgrade_ssd_version.py
- dist_test.sh
- dist_test_tracking.sh
- dist_train.sh
- slurm_test.sh
- slurm_test_tracking.sh
- slurm_train.sh
- test.py
- test_tracking.py
- train.py
README.md

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# WaterMask
![issues](https://img.shields.io/github/issues/LiamLian0727/WaterMask)
![forks](https://img.shields.io/github/forks/LiamLian0727/WaterMask?style=flat)
![stars](https://img.shields.io/github/stars/LiamLian0727/WaterMask?style=flat&color=red)
[![huggingface](https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Dataset-FFD21E)](https://huggingface.co/datasets/LiamLian0727/UIIS)
![license](https://img.shields.io/github/license/LiamLian0727/WaterMask)

---

### :speech_balloon: Updates

**[2026-03] Upgrade to MMDetection 3.0+**: We have upgraded the code from MMDetection 2.0 to **MMDetection 3.0+**! The main reasons for this upgrade are:
- MMDet 2.0 is outdated and no longer actively maintained
- MMDet 3.0 provides better performance, cleaner APIs, and improved modularity
- Better compatibility with latest PyTorch versions (2.0+) and CUDA
- Access to newer model architectures and training techniques

The original MMDet 2.0 version is still available on the [mmdet2 branch](https://github.com/LiamLian0727/WaterMask/tree/mmdet2).

**[2025-07] UIIS10K**:

[**UIIS10K**](https://github.com/LiamLian0727/UIIS10K/blob/main/README.md) upgrades the original UIIS dataset with **10 048 underwater images** annotated at **pixel level** for **10 object classes** (fish · reptiles · artiodactyla · mollusks · corals · plants · garbage · ruins · divers · robots).

- **Largest underwater benchmark** – more images and masks than any existing general underwater instance segmentation dataset.  
- **Scene diversity** – shallow & deep water, clear & turbid conditions, multiple resolutions, complex backgrounds, heavy occlusions.  
- **Crowded scenes** – 23 % of images contain ≥ 5 instances.  (up to 80+ per image).  
- **Multi-task ready** – pixel masks + bounding boxes support instance segmentation, object detection, and semantic segmentation.

---

This repo is the official implementation of "[WaterMask: Instance Segmentation for Underwater Imagery](https://openaccess.thecvf.com/content/ICCV2023/html/Lian_WaterMask_Instance_Segmentation_for_Underwater_Imagery_ICCV_2023_paper.html)". By Shijie Lian, Hua Li, Runmin Cong, Suqi Li, Wei Zhang, Sam Kwong, and has been accepted by ICCV2023! 🎉🎉🎉

### :rocket: Highlights:
- **UIIS dataset**: We construct the first general Underwater Image Instance Segmentation (UIIS) dataset containing 4,628 images for 7 categories with pixel-level annotations for underwater instance segmentation task.
  
  ![dataset img](https://raw.githubusercontent.com/LiamLian0727/WaterMask/mmdet3/figs/uiis_dataset.png)
- **SOTA performance**: We propose the first underwater instance segmentation model, WaterMask, as far as we know, which achieves good performance in qualitative and quantitative comparisons with natural image instance segmentation SOTA methods.
  
  ![framework_img](https://raw.githubusercontent.com/LiamLian0727/WaterMask/mmdet3/figs/framework.png)

## Installation

### Requirements
* Python 3.7+
* PyTorch 2.0+ (we use PyTorch 2.4.1)
* CUDA 12.1 or other version
* mmengine
* mmpretrain
* mmcv>=2.0.0
* [MMDetection](https://mmdetection.readthedocs.io/en/latest/get_started.html) 3.0+

### Environment Installation

**Step 0**: Download and install [Miniconda](https://docs.conda.io/projects/miniconda/en/latest/index.html) from the official website.

**Step 1**: Create a conda environment and activate it.

```shell
conda create -n watermask python=3.9 -y
conda activate watermask
```

**Step 2**: Install [PyTorch](https://pytorch.org/get-started/previous-versions/#v212). If you have experience with PyTorch and have already installed it, you can skip to the next section.

```shell
pip install torch==2.4.1 torchvision==0.19.1 --index-url https://download.pytorch.org/whl/cu121
```

**Step 3**: Install MMEngine, MMCV, and MMDetection using MIM.

```shell
pip install -U openmim
mim install mmengine
mim install "mmcv>=2.0.0"
mim install mmpretrain
mim install mmdet
```

### Datasets
Please create a data folder in your working directory and put the UIIS dataset in it for training or testing. UIIS is divided into two parts, with 3937 images for training and 691 images for testing.

data
      ├── UDW
      |   ├── annotations
      │   │   │   ├── train.json
      │   │   │   ├── val.json
      │   ├── train
      │   │   ├── L_1.jpg
      │   │   ├── ......
      │   ├── ......

you can get our UIIS dataset in [Hugging Face](https://huggingface.co/datasets/LiamLian0727/UIIS)、[Baidu Disk](https://pan.baidu.com/s/1owt1KZTnxzpSBiXxULtvCg?pwd=fiuk) (pwd:fiuk) or [Google Drive](https://drive.google.com/file/d/1MwGvsr2kJgKBGbU3zoZlXiqSSiP1ysGR/view?usp=sharing)

### Main Results
 We provide some results on **UIIS dataset** with pretrained models. These model are trained on an NVIDIA A5000 GPU. Note that all models and logs are available at [Baidu Netdisk](https://pan.baidu.com/s/1ZRgpcnaflNOmU_Zgfhv1mA?pwd=vhow) and [Google Drive](https://drive.google.com/file/d/1SlgkcOzYjKku15ldezNRe4O4OsqdprIB/view?usp=drive_link) link.

| model    |Schedule    | mAP        | AP50      |AP75      |   config |   download |
|:--------:|:----------:|:----------:|:---------:|:--------:|:--------:|:-------------:|
| WaterMask R101-FPN| 1x     | 25.6   | 41.7 |27.9 | [config](configs/_our_/water_r101_fpn_1x.py)  |   [log](https://pan.baidu.com/s/18feemQZP5wQMrmzRUmFy0g?pwd=j9pv) / [pth](https://pan.baidu.com/s/1EXjZAHrLbB9-mwIA6MeIbA?pwd=f6vd)
| WaterMask R50-FPN | 3x     | 26.4   | 43.6 |28.8   | [config](configs/_our_/water_r50_fpn_ms3x.py)   |   [log](https://pan.baidu.com/s/16_CQjHjeoooQEYasGnadWA?pwd=m4p5) / [pth](https://pan.baidu.com/s/1sNndq8LU4OtQXjpHcGNepg?pwd=3alo)
| WaterMask R101-FPN | 3x     | 27.2   | 43.7 |29.3  | [config](configs/_our_/water_r101_fpn_ms3x.py)   |   [log](https://pan.baidu.com/s/1JHxr1ybDjZ_c-olH7ugCFA?pwd=dgtp) / [pth](https://pan.baidu.com/s/1ciVAmXnfSQCckUufUd_wVA?pwd=7fzb)
| Cascade WaterMask R101-FPN| 3x| 27.1 | 42.9 |30.4  | [config](https://pan.baidu.com/s/1UW50jPH9nSeLPV5MjiKwSw?pwd=ouho)|[log](https://pan.baidu.com/s/1J3KkYkTQT62BMdrdYkH7-g?pwd=havk) / [pth](https://pan.baidu.com/s/12AE3oo09pPpAiW2PiztVhg?pwd=8nm2)

### Training
```
python tools/train.py configs/_our_/water_r50_fpn_1x.py --work-dir you_dir_to_save_logs_and_models
```

or

```
bash tools/dist_train.sh configs/_our_/water_r50_fpn_1x.py NUM_GPUS --work-dir you_dir_to_save_logs_and_models
```

### Test
```
python tools/test.py configs/_our_/water_r50_fpn_1x.py model_checkpoint_path --eval segm
```

or

```
bash tools/dist_test.sh configs/_our_/water_r50_fpn_1x.py model_checkpoint_path NUM_GPUS --eval segm
```

### Citation
If you find our repo useful for your research, please cite us:
```
@InProceedings{UIIS_Dataset_2023,
    author    = {Shijie Lian, Hua Li, Runmin Cong, Suqi Li, Wei Zhang, Sam Kwong},
    title     = {WaterMask: Instance Segmentation for Underwater Imagery},
    booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},
    month     = {October},
    year      = {2023},
    pages     = {1305-1315}
}

@article{UIIS10K_Dataset_2025,
    author    = {Hua Li, Shijie Lian, Zhiyuan Li, Runmin Cong, Chongyi Li},
    title     = {Taming SAM for Underwater Instance Segmentation and Beyond},
    year      = {2025},
    journal   = {arXiv preprint arXiv:2505.15581},
}
```

### License
This project is released under the [Apache 2.0 license](LICENSE).

### Acknowledgement
This software repository is implemented based on the [MMDetection](https://github.com/open-mmlab/mmdetection/tree/v2.25.1) framework. Thanks to them for their excellent work.