Natural-language retrieval of personal video archives through sparse-frame vision-language captioning and text-embedding search
Crisan, Paul Teodor
This work began outside any research setting. A small group of friends spent a stretch of student life travelling, going out, and taking part in the kinds of activities worth remembering. We captured almost all of it on camera, using phones, a drone, and an action camera, with the shared intent of one day turning the b...
- Spain
- English