Segment-Anything-finetune - GitLite

KO | EN

gitlite — search

#ai #typescript #golang #machine-learning #deep-learning #linux #docker #nodejs #macos #cli #vue #awesome-list #ios #open-source

Segment-Anything-finetune

(★ 156)

No description available.

File Explorer

Download Latest Version (.zip)

- 00000001.jpg
- .gitkeep
- .gitkeep
- .gitkeep
- learn_mask.py
- scores_debug.py
- - __init__.py
  - common.py
  - image_encoder.py
  - mask_decoder.py
  - prompt_encoder.py
  - sam.py
  - transformer.py
- - __init__.py
  - amg.py
  - onnx.py
  - transforms.py
- __init__.py
- automatic_mask_generator.py
- build_sam.py
- predictor.py
- convert_to_white_masks.py
- save_json.py
- save_labels.py
.gitignore
demo_app.py
LICENSE
README.md
requirements.txt
sam_eval.py
sam_finetune.py
sam_test.py

# Installation Guide

1. Get the code

git clone https://github.com/pg56714/Segment-Anything-finetune

Downloads the entire project code from GitHub to your computer.

cd Segment-Anything-finetune

Moves into the project folder you just downloaded.

2. Official Install Script

Easy Recommended

Prerequisites

Python 3 Python is required to use pip.

pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118

Installs the package published on PyPI directly — no need to clone the source.

✅ After installing, open a new terminal and run the program's version command (e.g. --version) to confirm it worked.

Pulled directly from this repo's README.

3. Python

Easy

Prerequisites

Git Needed to download the project code from GitHub.
Python 3 On Windows, be sure to check 'Add Python to PATH' during installation.

pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118

Installs the package published on PyPI directly — no need to clone the source.

pip install -r requirements.txt

Installs the Python libraries listed in requirements.txt (or similar).

✅ If it runs without errors and prints output in the terminal, it worked.

Pulled directly from this repo's README.

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# Segment-Anything-finetune

## Getting Started

```bash
conda create --name sam-finetune python=3.10 -y

conda activate sam-finetune

pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118

pip install -r requirements.txt
```

## Datasets Steps

1. Download from the [ViSha Dataset Source Link](https://erasernut.github.io/ViSha.html) or [SOBA_v2-Datasets](https://drive.google.com/drive/folders/1MKxyq3R6AUeyLai9i9XWzG2C_n5f0ppP) and place the files inside the `datasets` folder.

2. Run the `save_json.py` script in the `tool` folder to create `sam_train.json` and `sam_test.json` and place them inside the `datasets` folder.

3. Finish.

You can use the `save_labels.py` script in the `tool` folder to verify the labels.

Organized [SOBA_v2-Datasets](https://drive.google.com/drive/folders/1561wGAf0oik7C7__3byLHBNJOIadFuMw?usp=sharing) for use with the SAM model.

### Weights

Download the weights from the following links and save them in the `weights` directory.

[ViT-B](https://dl.fbaipublicfiles.com/segment_anything/sam_vit_b_01ec64.pth)

## Fine-tune

```bash
python sam_finetune.py
```

## Test

```bash
python sam_test.py
```

## Eval

```bash
python sam_eval.py
```

## Demo

```bash
python demo_app.py
```

## Source

[segment-anything](https://github.com/facebookresearch/segment-anything)

[Detect-AnyShadow](https://github.com/harrytea/Detect-AnyShadow)

[ViSha-Dataset-Source-Link](https://erasernut.github.io/ViSha.html)

[ViSha-Dataset-Link](https://drive.google.com/drive/folders/1Hp2mn_ui8I9GfxpXbLZ8zOvLlc_YJH4X)

[learn-how-to-fine-tune-the-segment-anything-model-sam](https://encord.com/blog/learn-how-to-fine-tune-the-segment-anything-model-sam/)

[fine-tune-the-segment-anything-model-sam-colab](https://colab.research.google.com/drive/1F6uRommb3GswcRlPZWpkAQRMVNdVH7Ww?usp=sharing#scrollTo=WRQ6yd_PM_B9)

[SOBA_v2-Datasets](https://drive.google.com/drive/folders/1MKxyq3R6AUeyLai9i9XWzG2C_n5f0ppP)

[SSIS](https://github.com/stevewongv/SSIS)