{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/adc1e9a402644b0f9ffeb3c7754e141a\" frameborder=\"0\" width=\"1662\" height=\"1246\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":1246,"width":1662,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":1246,"thumbnail_width":1662,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/adc1e9a402644b0f9ffeb3c7754e141a-0faeed3af7ee9d7b.gif","duration":556.965,"title":"Natural Language Semantic Video Search Engine","description":"This Loom demonstrates a semantic video search engine that finds specific moments in a large video library using natural language queries. It addresses the problem of time-consuming clip searching for video editors by indexing about 200 copyright-free stock videos downloaded from Pexabay and Pexels. The project uses OpenAI’s CLIP to embed images and text into a vector space, then FAISS to retrieve the closest matching frames sampled from videos every 2 or 3 seconds. Viewers see example queries like “a red car driving at night” and “a person opens a laptop,” along with an ingest option to add new videos to the library."}