# ArrowLuo/CLIP4Clip

An official implementation for "CLIP4Clip: An Empirical Study of CLIP for End to End Video Clip Retrieval"

Repository: https://github.com/ArrowLuo/CLIP4Clip
Canonical: https://ross.abutalabs.com/products/clip4clip
Homepage: https://arxiv.org/abs/2104.08860
Language: Python
License: MIT
License Family: permissive
Topics: video-text-retrieval, multimodal-learning, multimodality, multimodal, search, ranking, retrieval-model, retrieval, msrvtt, lsmdc, msvd, activitynet, didemo, video-clip-retrieval, clip
Last push: 2024-04-12T14:37:10+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 1968, "days_push": 873, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1031, forks 137 (observed 2026-08-28T04:03:17.983171+00:00)

## What it is
Official PyTorch implementation of the CLIP4Clip paper, a video-text retrieval model that transfers CLIP knowledge to end-to-end video clip retrieval. It benchmarks parameter-free, sequential, and tight similarity approaches on MSR-VTT, MSVD, LSMDC, ActivityNet, and DiDeMo.

## Use cases
- retrieve video clips from a text query
- search videos by natural language description
- reproduce CLIP4Clip paper results on MSR-VTT
- fine-tune CLIP for video-text retrieval
- benchmark video retrieval on MSVD, LSMDC, ActivityNet, DiDeMo
- build a multimodal video search prototype

## When to choose
- you need a proven baseline for text-to-video retrieval research
- you want to fine-tune CLIP on video-text datasets
- you need reproducible SOTA results on standard video retrieval benchmarks

## When to avoid
- you need a production-ready video search service with an API
- you work with images only rather than video
- you cannot access the required benchmark datasets (e.g., LSMDC requires permission)

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, search-engine, image-processing, video-processing, nlp
- domain: computer-vision, machine-learning
- platform: python
- tags: clip, video-text-retrieval, multimodal, research-code, msrvtt, video-clip-retrieval, search, natural-language-processing, video, gpu

## Member repositories
- ArrowLuo/CLIP4Clip (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:17.983171+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 2026-08-30T07:07:24.242900+00:00, confidence not recorded.
  - readme: https://github.com/ArrowLuo/CLIP4Clip (fetched 2026-08-28T04:03:17.983171+00:00, sha 95b20585d1c0)
  - homepage: https://arxiv.org/abs/2104.08860 (fetched 2026-08-29T13:06:56.732944+00:00, sha 2bb8fe641734)
  - site_page: https://info.arxiv.org/about/ourmembers.html (fetched 2026-08-29T13:06:56.745617+00:00, sha 47cbc55ff1de)
  - site_page: https://info.arxiv.org/about/donate.html (fetched 2026-08-29T13:06:56.742365+00:00, sha cca9c3a11c56)
  - site_page: https://info.arxiv.org/about (fetched 2026-08-29T13:06:56.747622+00:00, sha a1f16f915a9a)
  - site_page: https://info.arxiv.org/labs/index.html (fetched 2026-08-29T13:06:56.744044+00:00, sha b14a8d05a0ec)
- Data as of 2026-08-30T08:39:29.467469+00:00.