SentrySearch: Open-Source Semantic Video Search Tool Using Gemini Embedding 2
By
sohamrj
Sesame, salt, and substance. A flagship bake.
Summary
SentrySearch is an open-source tool that enables semantic search over video content. It works by splitting MP4 videos into overlapping chunks, embedding each chunk using either Google's Gemini Embedding API or a local Qwen3-VL model, and storing the vectors in a local ChromaDB database. Users can search by typing text queries, which are embedded into the same vector space and matched against stored video embeddings. The system then automatically trims and saves the top matching clip from the original video file.
Key quotes
· 4 pulledSentrySearch splits your mp4 videos into overlapping chunks, embeds each chunk as video using either Google's Gemini Embedding API or a local Qwen3-VL model, and stores the vectors in a local ChromaDB database.
When you search, your text query is embedded into the same vector space and matched against the stored video embeddings.
The top match is automatically trimmed from the original file and saved as a clip.
Semantic search over video footage. Type what you're looking for, get a trimmed clip back.
You might also wanna read
Google Releases Gemini Embedding 2: First Natively Multimodal Embedding Model
Google has released Gemini Embedding 2, its first natively multimodal embedding model that can map text, images, video, audio, and documents
Product Hunt·2mo ago
Edit Mind: A local-first video search tool that transcribes and indexes footage for natural language queries
Ilias, a content creator and software developer, built a local-first tool called "Edit Mind" that transcribes, analyzes frames (faces, objec
Product Hunt·18d ago