BIRADS-Classification-and-Feature-Recognition-of-Supernumerary-Mammary-Imaging

KO | EN

gitlite — search

#hacktoberfest #react #ai #llm #deep-learning #macos #nodejs #java #windows #awesome #cli #awesome-list #pytorch #mcp

BIRADS-Classification-and-Feature-Recognition-of-Supernumerary-Mammary-Imaging

(★ 10)

This is the individual project that won first prize in the national final of the Algorithm Challenge in the 6th Global Campus Artificial Intelligence Algorithm Elite Competition 2024, focusing on BIRADS classification and feature recognition based on ultrasound breast images.

BIRADS-Classification-and-Feature-Recognition-of-Supernumerary-Mammary-Imaging 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# 超声乳腺分类及 BI-RADS 特征检测（精简版）

这是一个精简后的 PyTorch 项目，包含 2 个统一入口：
- `train.py`：统一训练入口（`--task cls|feat`）
- `infer.py`：使用两个模型分别做分类与特征推理

## 项目结构

```text
.
├─datasets_core.py
├─train.py
├─infer.py
├─requirements.txt
├─.gitignore
├─LICENSE
└─data/                   # 本地数据，默认不上传
```

## 环境安装

推荐 Python `3.10+`

```bash
pip install -r requirements.txt
```

## 数据结构

### 分类数据

```text
data/乳腺分类训练数据集/
├─train/
│  ├─2类/images/
│  ├─3类/images/
│  ├─4A类/images/
│  ├─4B类/images/
│  ├─4C类/images/
│  └─5类/images/
└─test/
  └─(同上结构)
```

### 特征数据

```text
data/乳腺特征训练数据集/
├─train/
│  ├─images/
│  ├─boundary_labels/
│  ├─calcification_labels/
│  ├─direction_labels/
│  └─shape_labels/
└─test/
  └─(同上结构)
```

## 训练

```bash
python train.py --task cls --train-dir "data/乳腺分类训练数据集/train" --val-dir "data/乳腺分类训练数据集/test" --pretrained --epochs 20 --batch-size 16
python train.py --task feat --train-dir "data/乳腺特征训练数据集/train" --val-dir "data/乳腺特征训练数据集/test" --pretrained --epochs 20 --batch-size 16
```

常用参数：

```bash
python train.py --task cls --train-dir <分类训练目录> --val-dir <分类验证目录> --save-dir outputs
python train.py --task feat --train-dir <特征训练目录> --val-dir <特征验证目录> --save-dir outputs
```

## 推理

```bash
python infer.py --cls-weights "outputs/cls/best_cls_model.pth" --feat-weights "outputs/feat/best_feat_model.pth"
```

输出在 `outputs/predict/`：
- `cls_predictions.txt`
- `feat_predictions.txt`

## 上传到 GitHub

```bash
git init
git add .
git commit -m "refactor: simplify project structure"
git branch -M main
git remote add origin <你的仓库地址>
git push -u origin main
```

## 备注

- `data/`、权重、日志与输出目录已在 `.gitignore` 中排除。
- 训练脚本使用命令行参数，不再依赖硬编码绝对路径。
- 此项目所使用数据集为24年算法精英大赛超声乳腺数据集，只是为了提供一个思路，希望可以对你有所帮助
- 协作创作者： QinRui Zhu|zqr5  &  FeiYu Sun|QkkooQ

BIRADS-Classification-and-Feature-Recognition-of-Supernumerary-Mammary-Imaging

BIRADS-Classification-and-Feature-Recognition-of-Supernumerary-Mammary-Imaging 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

BIRADS-Classification-and-Feature-Recognition-of-Supernumerary-Mammary-Imaging

BIRADS-Classification-and-Feature-Recognition-of-Supernumerary-Mammary-Imaging 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README