Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Semantic search. Search local photos and videos through natural language. AI语义搜索本地素材。以图搜图、查找本地素材、根据文字描述匹配画面、视频帧搜索、根据画面描述搜索视频。
| Date | Stars |
|---|---|
| 2026-07-31 | 1928 |
| 2026-08-05 | 1929 |
| 2026-08-06 | 1929 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
# MaterialSearch [**中文**](./README_ZH.md) | [**English**](./README.md) Search local photos and videos through natural language. This repository now contains only the front-end code of the project. The core logic has been separated into a standalone pip package, which is hosted in another repository: [materialsearch-core](https://github.com/chn-lee-yumi/MaterialSearch-core). This structure makes version control and distribution easier, and also allows developers to use the core functions directly in their own projects. Some parts of the source code in this repository have been intentionally obfuscated. In the past, there were incidents where individuals maliciously removed or altered copyright and attribution information, which caused negative impact to the project. The obfuscation is solely for protecting authorship and legal integrity, and does **not** restrict any legitimate usage permitted under the GNU General Public License v3.0 (GPLv3). We kindly ask users to respect the original work and keep all relevant notices. If you make useful changes to the project (bug fixes, feature additions, etc.), you are welcome to contribute back through a pull request so that others can benefit. If you would like to contribute: - Frontend features: please submit a PR in this repository. - API-related requests: please submit an issue in this repository (the API implementation is not open-source and is maintained by the author). - Other features: please submit an issue here or a PR to the [materialsearch-core](https://github.com/chn-lee-yumi/MaterialSearch-core) repository. Note: AI-generated code contributions are not accepted. Please ensure that any submitted code is written and understood by yourself. ## Features - Text-based image search - Image-based image search - Text-based video search (provides matching video clips based on descriptions) - Image-based video search (searches for video segments based on screenshots) - Calculation of image-text similarity (provides a score, not very useful)  ## Deploy Instructions ### Windows Bundle Note: The minimum supported system is Windows 10. If you are still using Windows 7, please upgrade your system or switch to a newer computer. Tutorial video on Bilibili (Chinese): [Click here](https://www.bilibili.com/video/BV1SoqbB4Ehw/) User support QQ group (Chinese, community-driven): **1029566498** (The author has limited time, so the group is mainly for users to help each other.) First, download the bundle and extract it using **[7-Zip](https://www.7-zip.org/)**. Using other extraction tools may cause errors. There are two versions of the bundle: * `MaterialSearchWindows.7z`: does *not* include the model. Suitable for advanced users. * `MaterialSearchWindows_include_base_model.7z`: includes the base model (`OFA-Sys/chinese-clip-vit-base-patch16`) and works out of the box. **Recommended for most users.** Download links: * GitHub Release [https://github.com/chn-lee-yumi/MaterialSearch/releases/latest](https://github.com/chn-lee-yumi/MaterialSearch/releases/latest) * Quark Cloud [https://pan.quark.cn/s/ae137c439484](https://pan.quark.cn/s/ae137c439484) * Baidu Cloud [https://pan.baidu.com/s/1uQ8t-4mbYmcfi6FjwzdrrQ](https://pan.baidu.com/s/1uQ8t-4mbYmcfi6FjwzdrrQ) (code: CHNL) After extracting, please read the included `使用说明.txt` (Chinese) file. The bundle will automatically choose between integrated or dedicated GPU for acceleration. ### Deployment via Docker Supports `amd64` architectures. It includes the base models (`OFA-Sys/chinese-clip-vit-base-patch16`) and supports GPU acceleration. Image repositories: - [yumilee/materialsearch](https://hub.docker.com/r/yumilee/materialsearch) (DockerHub) - registry.cn-hongkong.aliyuncs.com/chn-lee-yumi/materialsearch (Aliyun, recommended for users in Mainland China) Before starting the image, you need to prepare: 1. The path to save the database 2. The scan paths on your local machine and th
Excerpt of 8,779 characters
Read on GitHub
329
9
3
2
1
1
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:0314f0d3f3b16aa5, llm:Description: 'Semantic search. Search local photos and videos through natural language... AI语义搜索本地素材。以图搜图、查找本地素材、根据文字描述匹配画面、视频帧搜索、根据画面描述搜索视频.' (semantic search of local images/videos, natural language, image-based search)
matched fp:0314f0d3f3b16aa5, llm:Description: 'Semantic search. Search local photos and videos through natural language... AI语义搜索本地素材。以图搜图、查找本地素材、根据文字描述匹配画面、视频帧搜索、根据画面描述搜索视频.' (semantic search of local images/videos, natural language, image-based search)
matched fp:0314f0d3f3b16aa5, llm:Description: 'Semantic search. Search local photos and videos through natural language... AI语义搜索本地素材。以图搜图、查找本地素材、根据文字描述匹配画面、视频帧搜索、根据画面描述搜索视频.' (semantic search of local images/videos, natural language, image-based search)
matched fp:0314f0d3f3b16aa5, llm:Description: 'Semantic search. Search local photos and videos through natural language... AI语义搜索本地素材。以图搜图、查找本地素材、根据文字描述匹配画面、视频帧搜索、根据画面描述搜索视频.' (semantic search of local images/videos, natural language, image-based search)