skills.rest
zilliztech
zilliztechOFFICIAL
Explore all skills by this author →
Website𝕏@zilliz_universeGitHub

video-search

Searches video content by text or visual via transcript and frame analysis.

3|2|Updated Jan 29, 2026
Search Tags:#semantic search#Whisper#transcripts#CLIP#frames#Milvus#video search
One-click install
npx skills add https://github.com/zilliztech/milvus-marketplace --skill video-search
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: video-search
Source: https://github.com/zilliztech/milvus-marketplace/tree/main/plugins/multimodal-retrieval/skills/video-search
Command: npx skills add https://github.com/zilliztech/milvus-marketplace --skill video-search

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This skill enables fast and accurate discovery of moments within video content by describing what you want to find and retrieving matching transcripts or visual scenes.

Core Features & Use Cases

  • Transcript-based search: locate spoken content in long videos using ASR transcripts.
  • Visual search: identify scenes by analyzing keyframes with CLIP-based similarity.
  • Unified workflow: handle meetings, tutorials, surveillance footage, lectures, movies, and sports by combining transcript and frame-based search.
  • Use Case: Imagine you have a training video and you want to jump to the moment where a specific term is mentioned or a particular visual cue appears.

Quick Start

To begin, describe the moment you want to locate or provide a reference frame or frame description, and the system will return matching video clips with timestamps.

Frequently Asked Questions about video-search

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I search video content by text or visual descriptions?▼

Video search locates specific moments by matching text queries against transcripts or visual queries against keyframes. Describe what you want to find—a spoken phrase or visual scene—and the system returns matching clips with timestamps using ASR transcripts and CLIP-based similarity matching.

Can I search video transcripts and frames at the same time?▼

Yes. The skill combines transcript-based search via ASR and visual search via CLIP keyframe analysis in a unified workflow, letting you find moments by spoken content or visual scenes within the same search.

What types of video content does this work with?▼

Transcript and frame-based search applies to meetings, tutorials, surveillance footage, lectures, films, and sports. Any video where you need to locate a specific moment by description or visual cue is suitable.

How does Milvus enable fast video moment retrieval?▼

Milvus performs fast similarity search on embeddings generated from transcripts and video frames, enabling rapid lookup of matching moments across large video libraries without scanning every second linearly.

What's the difference between searching transcripts versus keyframes?▼

Transcript search finds moments by matching spoken words extracted via ASR, while keyframe search identifies visual scenes by analyzing still frames with CLIP. Use transcripts for dialogue-heavy content and keyframes for visual-centric footage.

Do I need to provide my own video files or reference images?▼

Yes. You supply video content and either describe the moment you want to locate or provide a reference frame or frame description, and the system returns matching clips with timestamps.

Related Skills

Explore More Data & Analytics Skills→
JimLiuJimLiu
31

doubao-multimodal

Transcribe and analyze audio or video via the Doubao-Seed Ark endpoint.

Community
Advanced
jianshuojianshuo
114

wjs-segmenting-video

Segment long-form videos into themed clips using SRT timestamps and ffmpeg.

Community
Advanced
zilliztechzilliztech
3

image-search

Embed images with CLIP and retrieve visually similar items via a vector database.

Official
Advanced
PixelMLPixelML
2

agentic-video-memory

Ingest video files into a searchable database with transcription and embeddings.

Official
Advanced