[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-2434":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":16,"stars7d":17,"stars30d":18,"stars90d":16,"forks30d":16,"starsTrendScore":19,"compositeScore":20,"rankGlobal":10,"rankLanguage":10,"license":21,"archived":22,"fork":22,"defaultBranch":23,"hasWiki":24,"hasPages":22,"topics":25,"createdAt":10,"pushedAt":10,"updatedAt":38,"readmeContent":39,"aiSummary":40,"trendingCount":16,"starSnapshotCount":16,"syncStatus":41,"lastSyncTime":42,"discoverSource":43},2434,"HivisionIDPhotos","Zeyi-Lin\u002FHivisionIDPhotos","Zeyi-Lin","⚡️HivisionIDPhotos: a lightweight and efficient AI ID photos tools. 一个轻量级的AI证件照制作算法。","https:\u002F\u002Fmodelscope.cn\u002Fstudios\u002FSwanLab\u002FHivisionIDPhotos",null,"Python",21162,2416,86,96,0,15,70,7,45,"Apache License 2.0",false,"master",true,[26,27,28,29,30,31,32,33,34,35,36,37],"cnn","demo","docker","face-recognition","fastapi","gradio","idphoto","machine-learning","matting","mtcnn","tools","unet","2026-06-12 02:00:41","\u003Cdiv align=\"center\">\n\n\u003Cimg alt=\"hivision_logo\" src=\"assets\u002Fhivision_logo.png\" width=120 height=120>\n\u003Ch1>HivisionIDPhoto\u003C\u002Fh1>\n\n[English](README_EN.md) \u002F 中文 \u002F [日本語](README_JP.md) \u002F [한국어](README_KO.md)\n\n[![][release-shield]][release-link]\n[![][dockerhub-shield]][dockerhub-link]\n[![][github-stars-shield]][github-stars-link]\n[![][github-issues-shield]][github-issues-link]\n[![][github-contributors-shield]][github-contributors-link]\n[![][github-forks-shield]][github-forks-link]\n[![][license-shield]][license-link]  \n[![][wechat-shield]][wechat-link]\n[![][spaces-shield]][spaces-link]\n[![][swanhub-demo-shield]][swanhub-demo-link]\n[![][modelscope-shield]][modelscope-link]\n[![][modelers-shield]][modelers-link]\n[![][compshare-shield]][compshare-link]\n[![][atomgit-shield]][atomgit-link]\n\n[![][trendshift-shield]][trendshift-link]\n[![][hellogithub-shield]][hellogithub-link]\n\n\u003Cimg src=\"assets\u002FdemoImage.jpg\" width=900>\n\n\u003C\u002Fdiv>\n\n> **相关项目**：\n>\n> - [SwanLab](https:\u002F\u002Fgithub.com\u002FSwanHubX\u002FSwanLab)：一个开源、现代化设计的深度学习训练跟踪与可视化工具，同时支持云端\u002F离线使用，国内好用的Wandb平替；适配30+主流框架（PyTorch、HuggingFace Transformers、LLaMA Factory、Lightning等），欢迎使用！\n\n\n\u003Cbr>\n\n# 目录\n\n- [最近更新](#-最近更新)\n- [项目简介](#-项目简介)\n- [社区](#-社区)\n- [准备工作](#-准备工作)\n- [Demo启动](#-运行-gradio-demo)\n- [Python推理](#-python-推理)\n- [API服务部署](#️-部署-api-服务)\n- [Docker部署](#-docker-部署)\n- [联系我们](#-联系我们)\n- [FAQ](#faq)\n- [感谢支持](#-感谢支持)\n- [License](#-lincese)\n- [引用](#-引用)\n\n\u003Cbr>\n\n# 🤩 最近更新\n\n- 在线体验： [![Spaces](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002F🤗-Open%20in%20Spaces-blue)](https:\u002F\u002Fhuggingface.co\u002Fspaces\u002FTheEeeeLin\u002FHivisionIDPhotos)、[![][modelscope-shield]][modelscope-link]、[![][modelers-shield]][modelers-link]、[![][compshare-shield]][compshare-link]\n\n- 2024.11.20: Gradio Demo增加**打印排版**选项卡，支持六寸、五寸、A4、3R、4R五种排版尺寸\n- 2024.11.16: API接口增加美颜参数\n- 2024.09.25: 增加**五寸相纸**和**JPEG下载**选项｜默认照片下载支持300DPI\n- 2024.09.24: API接口增加base64图像传入选项 | Gradio Demo增加**排版照裁剪线**功能\n- 2024.09.22: Gradio Demo增加**野兽模式**，可设置内存加载策略 | API接口增加**dpi、face_alignment**参数\n- 2024.09.18: Gradio Demo增加**分享模版照**功能、增加**美式证件照**背景选项\n- 2024.09.17: Gradio Demo增加**自定义底色-HEX输入**功能 | **（社区贡献）C++版本** - [HivisionIDPhotos-cpp](https:\u002F\u002Fgithub.com\u002Fzjkhahah\u002FHivisionIDPhotos-cpp) 贡献 by [zjkhahah](https:\u002F\u002Fgithub.com\u002Fzjkhahah)\n- 2024.09.16: Gradio Demo增加**人脸旋转对齐**功能，自定义尺寸输入支持**毫米**单位\n\n\u003Cbr>\n\n# 项目简介\n\n> 🚀 谢谢你对我们的工作感兴趣。您可能还想查看我们在图像领域的其他成果，欢迎来信:zeyi.lin@swanhub.co.\n\nHivisionIDPhoto 旨在开发一种实用、系统性的证件照智能制作算法。\n\n它利用一套完善的AI模型工作流程，实现对多种用户拍照场景的识别、抠图与证件照生成。\n\n**HivisionIDPhoto 可以做到：**\n\n1. 轻量级抠图（纯离线，仅需 **CPU** 即可快速推理）\n2. 根据不同尺寸规格生成不同的标准证件照、六寸排版照\n3. 支持 纯离线 或 端云 推理\n4. 美颜\n5. 智能换正装（waiting）\n\n\u003Cdiv align=\"center\">\n\u003Cimg src=\"assets\u002Fdemo.png\" width=900>\n\u003C\u002Fdiv>\n\n---\n\n如果 HivisionIDPhoto 对你有帮助，请 star 这个 repo 或推荐给你的朋友，解决证件照应急制作问题！\n\n\u003Cbr>\n\n# 🏠 社区\n\n我们分享了一些由社区构建的HivisionIDPhotos的有趣应用和扩展：\n\n| [HivisionIDPhotos-ComfyUI][community-hivision-comfyui] | [HivisionIDPhotos-wechat-weapp][community-hivision-wechat] |\n| :----------------------------------------------------------------------------------------------------------------------------------------------: | :------------------------------------------------------------------------------------------------------------------------------: |\n| \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FAIFSH\u002FHivisionIDPhotos-ComfyUI\"> \u003Cimg src=\"assets\u002Fcomfyui.png\" width=\"900\" alt=\"ComfyUI workflow\"> \u003C\u002Fa>  | \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fno1xuan\u002FHivisionIDPhotos-wechat-weapp\"> \u003Cimg src=\"assets\u002Fcommunity-wechat-miniprogram.png\" width=\"900\" alt=\"ComfyUI workflow\"> \u003C\u002Fa>  |\n|ComfyUI证件照处理工作流 | 证件照微信小程序（JAVA后端+原生前端） |\n\n| [HivisionIDPhotos-Uniapp][community-hivision-uniapp] | [HivisionIDPhotos-web](https:\u002F\u002Fgithub.com\u002Fjkm199\u002FHivisionIDPhotos-web)|\n| :------------------------------------------------------------------------------------------------------------------------------: | :------------------------------------------------------------------------------------------------------------------------------: |\n| \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fsoulerror\u002FHivisionIDPhotos-Uniapp\"> \u003Cimg src=\"assets\u002Fcommunity-uniapp-wechat-miniprogram.png\" width=\"900\" alt=\"HivisionIDPhotos-uniapp\"> \u003C\u002Fa>  | \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fjkm199\u002FHivisionIDPhotos-web\"> \u003Cimg src=\"assets\u002Fcommunity-web.png\" width=\"900\" alt=\"HivisionIDPhotos-uniapp\"> \u003C\u002Fa>  |\n| 证件照微信小程序（uniapp）| 证件照应用网页版 |\n\n\n- [HivisionIDPhotos-cpp](https:\u002F\u002Fgithub.com\u002Fzjkhahah\u002FHivisionIDPhotos-cpp): HivisionIDphotos C++版本，由 [zjkhahah](https:\u002F\u002Fgithub.com\u002Fzjkhahah) 构建\n- [ai-idphoto](https:\u002F\u002Fgithub.com\u002Fwmlcjj\u002Fai-idphoto): [HivisionIDPhotos-wechat-weapp](https:\u002F\u002Fgithub.com\u002Fno1xuan\u002FHivisionIDPhotos-wechat-weapp) 的uniapp多端兼容版，由 [wmlcjj](https:\u002F\u002Fgithub.com\u002Fwmlcjj) 贡献\n- [HivisionIDPhotos-uniapp-WeChat-gpto1](https:\u002F\u002Fgithub.com\u002Fjkm199\u002FHivisionIDPhotos-uniapp-WeChat-gpto1\u002F): 由gpt-o1辅助完成开发的证件照微信小程序，由 [jkm199](https:\u002F\u002Fgithub.com\u002Fjkm199) 贡献\n- [HivisionIDPhotos-windows-GUI](https:\u002F\u002Fgithub.com\u002Fzhaoyun0071\u002FHivisionIDPhotos-windows-GUI)：Windows客户端应用，由 [zhaoyun0071](https:\u002F\u002Fgithub.com\u002Fzhaoyun0071) 构建\n- [HivisionIDPhotos-NAS](https:\u002F\u002Fgithub.com\u002FONG-Leo\u002FHivisionIDPhotos-NAS): 群晖NAS部署中文教程，由 [ONG-Leo](https:\u002F\u002Fgithub.com\u002FONG-Leo) 贡献\n\n\n\u003Cbr>\n\n# 🔧 准备工作\n\n环境安装与依赖：\n- Python >= 3.7（项目主要测试在 python 3.10）\n- OS: Linux, Windows, MacOS\n\n## 1. 克隆项目\n\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002FZeyi-Lin\u002FHivisionIDPhotos.git\ncd  HivisionIDPhotos\n```\n\n## 2. 安装依赖环境\n\n> 建议 conda 创建一个 python3.10 虚拟环境后，执行以下命令\n\n```bash\npip install -r requirements.txt\npip install -r requirements-app.txt\n```\n\n## 3. 下载人像抠图模型权重文件\n\n**方式一：脚本下载**\n\n```bash\npython scripts\u002Fdownload_model.py --models all\n# 如需指定下载某个模型\n# python scripts\u002Fdownload_model.py --models modnet_photographic_portrait_matting\n```\n\n**方式二：直接下载**\n\n模型均存到项目的`hivision\u002Fcreator\u002Fweights`目录下：\n\n| 人像抠图模型 | 介绍 | 下载 |\n| -- | -- | -- |\n| MODNet | [MODNet](https:\u002F\u002Fgithub.com\u002FZHKKKe\u002FMODNet)官方权重 | [下载](https:\u002F\u002Fgithub.com\u002FZeyi-Lin\u002FHivisionIDPhotos\u002Freleases\u002Fdownload\u002Fpretrained-model\u002Fmodnet_photographic_portrait_matting.onnx)(24.7MB)|\n| hivision_modnet | 对纯色换底适配性更好的抠图模型 | [下载](https:\u002F\u002Fgithub.com\u002FZeyi-Lin\u002FHivisionIDPhotos\u002Freleases\u002Fdownload\u002Fpretrained-model\u002Fhivision_modnet.onnx)(24.7MB) |\n| rmbg-1.4 | [BRIA AI](https:\u002F\u002Fhuggingface.co\u002Fbriaai\u002FRMBG-1.4) 开源的抠图模型 | [下载](https:\u002F\u002Fhuggingface.co\u002Fbriaai\u002FRMBG-1.4\u002Fresolve\u002Fmain\u002Fonnx\u002Fmodel.onnx?download=true)(176.2MB)后重命名为`rmbg-1.4.onnx` |\n| birefnet-v1-lite | [ZhengPeng7](https:\u002F\u002Fgithub.com\u002FZhengPeng7\u002FBiRefNet) 开源的抠图模型，拥有最好的分割精度 | [下载](https:\u002F\u002Fgithub.com\u002FZhengPeng7\u002FBiRefNet\u002Freleases\u002Fdownload\u002Fv1\u002FBiRefNet-general-bb_swin_v1_tiny-epoch_232.onnx)(224MB)后重命名为`birefnet-v1-lite.onnx` |\n\n> 如果下载网速不顺利：前往[SwanHub](https:\u002F\u002Fswanhub.co\u002FZeYiLin\u002FHivisionIDPhotos_models\u002Ftree\u002Fmain)下载。\n\n\n## 4. 人脸检测模型配置（可选）\n\n| 拓展人脸检测模型 | 介绍 | 使用文档 |\n| -- | -- | -- |\n| MTCNN | **离线**人脸检测模型，高性能CPU推理（毫秒级），为默认模型，检测精度较低 | Clone此项目后直接使用 |\n| RetinaFace | **离线**人脸检测模型，CPU推理速度中等（秒级），精度较高| [下载](https:\u002F\u002Fgithub.com\u002FZeyi-Lin\u002FHivisionIDPhotos\u002Freleases\u002Fdownload\u002Fpretrained-model\u002Fretinaface-resnet50.onnx)后放到`hivision\u002Fcreator\u002Fretinaface\u002Fweights`目录下 |\n| Face++ | 旷视推出的在线人脸检测API，检测精度较高，[官方文档](https:\u002F\u002Fconsole.faceplusplus.com.cn\u002Fdocuments\u002F4888373) | [使用文档](docs\u002Fface++_CN.md)|\n\n## 5. 性能参考\n\n> 测试环境为Mac M1 Max 64GB，非GPU加速，测试图片分辨率为 512x715(1) 与 764×1146(2)。\n\n| 模型组合 | 内存占用 | 推理时长(1) | 推理时长(2) |\n| -- | -- | -- | -- |\n| MODNet + mtcnn | 410MB | 0.207s | 0.246s |\n| MODNet + retinaface | 405MB | 0.571s | 0.971s |\n| birefnet-v1-lite + retinaface | 6.20GB | 7.063s | 7.128s |\n\n## 6. GPU推理加速（可选）\n\n在当前版本，可被英伟达GPU加速的模型为`birefnet-v1-lite`，并请确保你有16GB左右的显存。\n\n如需使用英伟达GPU加速推理，在确保你已经安装[CUDA](https:\u002F\u002Fdeveloper.nvidia.com\u002Fcuda-downloads)与[cuDNN](https:\u002F\u002Fdeveloper.nvidia.com\u002Fcudnn)后，根据[onnxruntime-gpu文档](https:\u002F\u002Fonnxruntime.ai\u002Fdocs\u002Fexecution-providers\u002FCUDA-ExecutionProvider.html#cuda-12x)找到对应的`onnxruntime-gpu`版本安装，以及根据[pytorch官网](https:\u002F\u002Fpytorch.org\u002Fget-started\u002Flocally\u002F)找到对应的`torch`版本安装。\n\n```bash\n# 假如你的电脑安装的是CUDA 12.x, cuDNN 8\n# 安装torch是可选的，如果你始终配置不好cuDNN，那么试试安装torch\npip install onnxruntime-gpu==1.18.0\npip install torch --index-url https:\u002F\u002Fdownload.pytorch.org\u002Fwhl\u002Fcu121\n```\n\n完成安装后，调用`birefnet-v1-lite`模型即可利用GPU加速推理。\n\n> TIPS: CUDA 支持向下兼容。比如你的 CUDA 版本为 12.6，`torch` 官方目前支持的最高版本为 12.4（\u003C12.6），`torch`仍可以正常使用CUDA。\n\n\u003Cbr>\n\n# ⚡️ 运行 Gradio Demo\n\n```bash\npython app.py\n```\n\n运行程序将生成一个本地 Web 页面，在页面中可完成证件照的操作与交互。\n\n\u003Cimg src=\"assets\u002Fharry.png\" width=900>\n\n\u003Cbr>\n\n# 🚀 Python 推理\n\n核心参数：\n\n- `-i`: 输入图像路径\n- `-o`: 保存图像路径\n- `-t`: 推理类型，有idphoto、human_matting、add_background、generate_layout_photos可选\n- `--matting_model`: 人像抠图模型权重选择\n- `--face_detect_model`: 人脸检测模型选择\n\n更多参数可通过`python inference.py --help`查看\n\n## 1. 证件照制作\n\n输入 1 张照片，获得 1 张标准证件照和 1 张高清证件照的 4 通道透明 png\n\n```python\npython inference.py -i demo\u002Fimages\u002Ftest0.jpg -o .\u002Fidphoto.png --height 413 --width 295\n```\n\n## 2. 人像抠图\n\n输入 1 张照片，获得 1张 4 通道透明 png\n\n```python\npython inference.py -t human_matting -i demo\u002Fimages\u002Ftest0.jpg -o .\u002Fidphoto_matting.png --matting_model hivision_modnet\n```\n\n## 3. 透明图增加底色\n\n输入 1 张 4 通道透明 png，获得 1 张增加了底色的 3通道图像\n\n```python\npython inference.py -t add_background -i .\u002Fidphoto.png -o .\u002Fidphoto_ab.jpg  -c 4f83ce -k 30 -r 1\n```\n\n## 4. 得到六寸排版照\n\n输入 1 张 3 通道照片，获得 1 张六寸排版照\n\n```python\npython inference.py -t generate_layout_photos -i .\u002Fidphoto_ab.jpg -o .\u002Fidphoto_layout.jpg  --height 413 --width 295 -k 200\n```\n\n## 5. 证件照裁剪\n\n输入 1 张 4 通道照片（抠图好的图像），获得 1 张标准证件照和 1 张高清证件照的 4 通道透明 png\n\n```python\npython inference.py -t idphoto_crop -i .\u002Fidphoto_matting.png -o .\u002Fidphoto_crop.png --height 413 --width 295\n```\n\n\n\u003Cbr>\n\n# ⚡️ 部署 API 服务\n\n## 启动后端\n\n```\npython deploy_api.py\n```\n\n## 请求 API 服务\n\n详细请求方式请参考 [API 文档](docs\u002Fapi_CN.md)，包含以下请求示例：\n- [cURL](docs\u002Fapi_CN.md#curl-请求示例)\n- [Python](docs\u002Fapi_CN.md#python-请求示例)\n\n\u003Cbr>\n\n# 🐳 Docker 部署\n\n## 1. 拉取或构建镜像\n\n> 以下方式三选一\n\n**方式一：拉取最新镜像：**\n\n```bash\ndocker pull linzeyi\u002Fhivision_idphotos\n```\n\n**方式二：Dockrfile 直接构建镜像：**\n\n在确保将至少一个[抠图模型权重文件](#3-下载权重文件)放到`hivision\u002Fcreator\u002Fweights`下后，在项目根目录执行：\n\n```bash\ndocker build -t linzeyi\u002Fhivision_idphotos .\n```\n\n**方式三：Docker compose 构建：**\n\n在确保将至少一个[抠图模型权重文件](#3-下载权重文件)放到`hivision\u002Fcreator\u002Fweights`下后，在项目根目录下执行：\n\n```bash\ndocker compose build\n```\n\n## 2. 运行服务\n\n**启动 Gradio Demo 服务**\n\n运行下面的命令，在你的本地访问 [http:\u002F\u002F127.0.0.1:7860](http:\u002F\u002F127.0.0.1:7860\u002F) 即可使用。\n\n```bash\ndocker run -d -p 7860:7860 linzeyi\u002Fhivision_idphotos\n```\n\n**启动 API 后端服务**\n\n```bash\ndocker run -d -p 8080:8080 linzeyi\u002Fhivision_idphotos python3 deploy_api.py\n```\n\n**两个服务同时启动**\n\n```bash\ndocker compose up -d\n```\n\n## 环境变量\n\n本项目提供了一些额外的配置项，使用环境变量进行设置：\n\n| 环境变量 | 类型\t| 描述 | 示例 |\n|--|--|--|--|\n| FACE_PLUS_API_KEY\t | 可选\t| 这是你在 Face++ 控制台申请的 API 密钥\t | `7-fZStDJ····` |\n| FACE_PLUS_API_SECRET\t | 可选\t| Face++ API密钥对应的Secret | `VTee824E····` |\n| RUN_MODE | 可选 | 运行模式，可选值为`beast`(野兽模式)。野兽模式下人脸检测和抠图模型将不释放内存，从而获得更快的二次推理速度。建议内存16GB以上尝试。 | `beast` |\n| DEFAULT_LANG | 可选 | Gradio Demo启动时的默认语言| `en` |\n\ndocker使用环境变量示例：\n```bash\ndocker run  -d -p 7860:7860 \\\n    -e FACE_PLUS_API_KEY=7-fZStDJ···· \\\n    -e FACE_PLUS_API_SECRET=VTee824E···· \\\n    -e RUN_MODE=beast \\\n    -e DEFAULT_LANG=en \\\n    linzeyi\u002Fhivision_idphotos  \n```\n\n\u003Cbr>\n\n# FAQ\n\n## 1. 如何修改预设尺寸和颜色？\n\n- 尺寸：修改[size_list_CN.csv](demo\u002Fassets\u002Fsize_list_CN.csv)后再次运行 `app.py` 即可，其中第一列为尺寸名，第二列为高度，第三列为宽度。\n- 颜色：修改[color_list_CN.csv](demo\u002Fassets\u002Fcolor_list_CN.csv)后再次运行 `app.py` 即可，其中第一列为颜色名，第二列为Hex值。\n\n## 2. 如何修改水印字体？\n\n1. 将字体文件放到`hivision\u002Fplugin\u002Ffont`文件夹下\n2. 修改`hivision\u002Fplugin\u002Fwatermark.py`的`font_file`参数值为字体文件名\n\n## 3. 如何添加社交媒体模板照？\n\n1. 将模板图片放到`hivision\u002Fplugin\u002Ftemplate\u002Fassets`文件夹下。模板图片是一个4通道的透明png。\n2. 在`hivision\u002Fplugin\u002Ftemplate\u002Fassets\u002Ftemplate_config.json`文件中添加最新的模板信息，其中`width`为模板图宽度(px)，`height`为模板图高度(px)，`anchor_points`为模板中透明区域的四个角的坐标(px)；`rotation`为透明区域相对于垂直方向的旋转角度，>0为逆时针，\u003C0为顺时针。\n3. 在`demo\u002Fprocessor.py`的`_generate_image_template`函数中的`TEMPLATE_NAME_LIST`变量添加最新的模板名\n\n\u003Cimg src=\"assets\u002Fsocial_template.png\" width=\"500\">\n\n## 4. 如何修改Gradio Demo的顶部导航栏？\n\n- 修改`demo\u002Fassets\u002Ftitle.md`\n\n## 5. 如何添加\u002F修改「打印排版」中的尺寸？\n\n- 修改`demo\u002Flocales.py`中的`print_switch`字典，添加\u002F修改新的尺寸名称和尺寸参数，然后重新运行`python app.py`\n\n\u003Cbr>\n\n# 📧 联系我们\n\n如果您有任何问题，请发邮件至 zeyi.lin@swanhub.co\n\n\u003Cbr>\n\n# 🙏 感谢支持\n\n[![Stargazers repo roster for @Zeyi-Lin\u002FHivisionIDPhotos](https:\u002F\u002Freporoster.com\u002Fstars\u002FZeyi-Lin\u002FHivisionIDPhotos)](https:\u002F\u002Fgithub.com\u002FZeyi-Lin\u002FHivisionIDPhotos\u002Fstargazers)\n\n[![Forkers repo roster for @Zeyi-Lin\u002FHivisionIDPhotos](https:\u002F\u002Freporoster.com\u002Fforks\u002FZeyi-Lin\u002FHivisionIDPhotos)](https:\u002F\u002Fgithub.com\u002FZeyi-Lin\u002FHivisionIDPhotos\u002Fnetwork\u002Fmembers)\n\n[![Star History Chart](https:\u002F\u002Fapi.star-history.com\u002Fsvg?repos=Zeyi-Lin\u002FHivisionIDPhotos&type=Date)](https:\u002F\u002Fstar-history.com\u002F#Zeyi-Lin\u002FHivisionIDPhotos&Date)\n\n贡献者们：\n\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FZeyi-Lin\u002FHivisionIDPhotos\u002Fgraphs\u002Fcontributors\">\n  \u003Cimg src=\"https:\u002F\u002Fcontrib.rocks\u002Fimage?repo=Zeyi-Lin\u002FHivisionIDPhotos\" \u002F>\n\u003C\u002Fa>\n\n[Zeyi-Lin](https:\u002F\u002Fgithub.com\u002FZeyi-Lin)、[SAKURA-CAT](https:\u002F\u002Fgithub.com\u002FSAKURA-CAT)、[Feudalman](https:\u002F\u002Fgithub.com\u002FFeudalman)、[swpfY](https:\u002F\u002Fgithub.com\u002FswpfY)、[Kaikaikaifang](https:\u002F\u002Fgithub.com\u002FKaikaikaifang)、[ShaohonChen](https:\u002F\u002Fgithub.com\u002FShaohonChen)、[KashiwaByte](https:\u002F\u002Fgithub.com\u002FKashiwaByte)\n\n\u003Cbr>\n\n# 📜 Lincese\n\nThis repository is licensed under the [Apache-2.0 License](LICENSE).\n\n\u003Cbr>\n\n# 📚 引用\n\n如果您在研究或项目中使用了HivisionIDPhotos，请考虑引用我们的工作。您可以使用以下BibTeX条目：\n\n```bibtex\n@misc{hivisionidphotos,\n      title={{HivisionIDPhotos: A Lightweight and Efficient AI ID Photos Tool}},\n      author={Zeyi Lin and SwanLab Team},\n      year={2024},\n      publisher={GitHub},\n      url = {\\url{https:\u002F\u002Fgithub.com\u002FZeyi-Lin\u002FHivisionIDPhotos}},\n}\n```\n\n\n\n\n[github-stars-shield]: https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fzeyi-lin\u002Fhivisionidphotos?color=ffcb47&labelColor=black&style=flat-square\n[github-stars-link]: https:\u002F\u002Fgithub.com\u002Fzeyi-lin\u002Fhivisionidphotos\u002Fstargazers\n\n[swanhub-demo-shield]: https:\u002F\u002Fswanhub.co\u002Fgit\u002Frepo\u002FSwanHub%2FAuto-README\u002Ffile\u002Fpreview?ref=main&path=swanhub.svg\n[swanhub-demo-link]: https:\u002F\u002Fswanhub.co\u002FZeYiLin\u002FHivisionIDPhotos\u002Fdemo\n\n[spaces-shield]: https:\u002F\u002Fimg.shields.io\u002Fbadge\u002F🤗-Open%20in%20Spaces-blue\n[spaces-link]: https:\u002F\u002Fhuggingface.co\u002Fspaces\u002FTheEeeeLin\u002FHivisionIDPhotos\n\n\u003C!-- 微信群链接 -->\n[wechat-shield]: https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FWeChat-微信-4cb55e\n[wechat-link]: https:\u002F\u002Fdocs.qq.com\u002Fdoc\u002FDUkpBdk90eWZFS2JW\n\n\u003C!-- Github Release -->\n[release-shield]: https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fv\u002Frelease\u002Fzeyi-lin\u002Fhivisionidphotos?color=369eff&labelColor=black&logo=github&style=flat-square\n[release-link]: https:\u002F\u002Fgithub.com\u002Fzeyi-lin\u002Fhivisionidphotos\u002Freleases\n\n[license-shield]: https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Flicense-apache%202.0-white?labelColor=black&style=flat-square\n[license-link]: https:\u002F\u002Fgithub.com\u002FZeyi-Lin\u002FHivisionIDPhotos\u002Fblob\u002Fmaster\u002FLICENSE\n\n[github-issues-shield]: https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fissues\u002Fzeyi-lin\u002Fhivisionidphotos?color=ff80eb&labelColor=black&style=flat-square\n[github-issues-link]: https:\u002F\u002Fgithub.com\u002Fzeyi-lin\u002Fhivisionidphotos\u002Fissues\n\n[dockerhub-shield]: https:\u002F\u002Fimg.shields.io\u002Fdocker\u002Fv\u002Flinzeyi\u002Fhivision_idphotos?color=369eff&label=docker&labelColor=black&logoColor=white&style=flat-square\n[dockerhub-link]: https:\u002F\u002Fhub.docker.com\u002Fr\u002Flinzeyi\u002Fhivision_idphotos\u002Ftags\n\n[trendshift-shield]: https:\u002F\u002Ftrendshift.io\u002Fapi\u002Fbadge\u002Frepositories\u002F11622\n[trendshift-link]: https:\u002F\u002Ftrendshift.io\u002Frepositories\u002F11622\n\n[hellogithub-shield]: https:\u002F\u002Fabroad.hellogithub.com\u002Fv1\u002Fwidgets\u002Frecommend.svg?rid=8ea1457289fb4062ba661e5299e733d6&claim_uid=Oh5UaGjfrblg0yZ\n[hellogithub-link]: https:\u002F\u002Fhellogithub.com\u002Frepository\u002F8ea1457289fb4062ba661e5299e733d6\n\n[github-contributors-shield]: https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fcontributors\u002Fzeyi-lin\u002Fhivisionidphotos?color=c4f042&labelColor=black&style=flat-square\n[github-contributors-link]: https:\u002F\u002Fgithub.com\u002Fzeyi-lin\u002Fhivisionidphotos\u002Fgraphs\u002Fcontributors\n\n[github-forks-shield]: https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fforks\u002Fzeyi-lin\u002Fhivisionidphotos?color=8ae8ff&labelColor=black&style=flat-square\n[github-forks-link]: https:\u002F\u002Fgithub.com\u002Fzeyi-lin\u002Fhivisionidphotos\u002Fnetwork\u002Fmembers\n\n[modelscope-shield]: https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FDemo_on_ModelScope-purple?logo=data:image\u002Fsvg+xml;base64,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&labelColor=white\n[modelscope-link]: https:\u002F\u002Fmodelscope.cn\u002Fstudios\u002FSwanLab\u002FHivisionIDPhotos\n\n[modelers-shield]: https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FDemo_on_Modelers-c42a2a?logo=data:image\u002Fsvg+xml;base64,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&labelColor=white\n[modelers-link]: https:\u002F\u002Fmodelers.cn\u002Fspaces\u002FSwanLab\u002FHivisionIDPhotos\n\n[compshare-shield]: https:\u002F\u002Fwww-s.ucloud.cn\u002F2025\u002F02\u002Fdbef8b07ea3d316006d9c22765c3cd53_1740104342584.svg\n[compshare-link]: https:\u002F\u002Fwww.compshare.cn\u002Fimages-detail?ImageID=compshareImage-17jacgm4ju16&ytag=HG_GPU_HivisionIDPhotos\n\n[atomgit-shield]: https:\u002F\u002Fatomgit.com\u002FZeYiLin\u002FHivisionIDPhotos\u002Fstar\u002Fbadge.svg\n[atomgit-link]: https:\u002F\u002Fatomgit.com\u002FZeYiLin\u002FHivisionIDPhotos\n\n\u003C!-- 社区项目链接 -->\n[community-hivision-comfyui]: https:\u002F\u002Fgithub.com\u002FAIFSH\u002FHivisionIDPhotos-ComfyUI\n[community-hivision-wechat]: https:\u002F\u002Fgithub.com\u002Fno1xuan\u002FHivisionIDPhotos-wechat-weapp\n[community-hivision-uniapp]: https:\u002F\u002Fgithub.com\u002Fsoulerror\u002FHivisionIDPhotos-Uniapp\n[community-hivision-cpp]: https:\u002F\u002Fgithub.com\u002Fzjkhahah\u002FHivisionIDPhotos-cpp\n[community-hivision-windows-gui]: https:\u002F\u002Fgithub.com\u002Fzhaoyun0071\u002FHivisionIDPhotos-windows-GUI\n[community-hivision-nas]: https:\u002F\u002Fgithub.com\u002FONG-Leo\u002FHivisionIDPhotos-NAS\n","HivisionIDPhotos 是一个轻量级且高效的AI证件照制作工具。该项目利用先进的CNN、MTCNN和UNet等深度学习技术，实现对用户上传照片的智能识别、抠图及证件照生成。其核心功能包括轻量级抠图（仅需CPU即可快速处理）、支持多种标准证件照尺寸规格的自定义生成、以及美颜效果调整。此外，它还提供了Gradio界面供用户轻松体验，并支持Docker部署与API服务集成。HivisionIDPhotos非常适合需要快速生成高质量证件照的个人用户或小型企业使用，在保证高效性的同时也兼顾了易用性和灵活性。",2,"2026-06-11 02:49:55","top_language"]