# chn-lee-yumi/MaterialSearch

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。

Repository: https://github.com/chn-lee-yumi/MaterialSearch
Canonical: https://ross.abutalabs.com/products/materialsearch
Homepage: http://materialsearch.gcc.ac.cn/
Language: HTML
License: GPL-3.0
License Family: copyleft
Last push: 2026-05-23T06:27:12+00:00

## Health v2 (maintenance only)
Score: 73/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 83, release rhythm 51, longevity 90
- inputs: {"age_days": 1266, "days_push": 102, "days_rel": 167, "gap_med": 92, "n_releases_24m": 2}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1949, forks 217 (observed 2026-08-28T04:05:58.093352+00:00)

## What it is
MaterialSearch is a self-hosted semantic search tool that indexes local photos and videos using a CLIP multimodal model, letting users find media via natural-language text or by example images. It runs fully offline with a browser-based front-end and a Python core available as a pip package.

## Use cases
- find local photos by describing them in text
- reverse image search across my local photo library
- search videos by describing the scene
- find which video a screenshot came from
- locate video clips matching a text description
- index my NAS photo and video collection for semantic search

## When to choose
- you have a large local media library and no organized tags or filenames
- you need fully offline, privacy-preserving media search
- you want both text-to-image and image-to-video reverse search
- you want a Docker or Windows bundle deployment on a PC or NAS

## When to avoid
- you need cloud-scale or multi-user hosted search
- you require OCR or text-in-image search rather than visual semantics
- you need a fully open-source API implementation (the API layer is not open-source)
- you cannot run GPU or reasonably fast CPU inference for indexing

## Facets
- artifact type: application
- maturity: active
- function: search-engine, image-processing, video-processing, machine-learning, http-server
- domain: image-processing, artificial-intelligence, self-hosted, media
- platform: windows, self-hosted, python
- tags: semantic-search, clip, reverse-image-search, local-photos, video-frame-search, privacy-first, multimodal, video, search, linux, docker, web-server

## Member repositories
- chn-lee-yumi/MaterialSearch (main) score 73

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:58.093352+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:06:54.199603+00:00, confidence not recorded.
  - readme: https://github.com/chn-lee-yumi/MaterialSearch (fetched 2026-08-28T04:05:58.093352+00:00, sha efd18ab3ea02)
  - homepage: http://materialsearch.gcc.ac.cn/ (fetched 2026-08-29T10:46:44.500053+00:00, sha 461950532670)
- Data as of 2026-08-30T08:39:29.467469+00:00.