videoduplicatefinder

Why we included this project

The usual way to find duplicate videos is hashing, which only catches files that are byte-for-byte identical, so a re-encoded or watermarked copy slips right past. Video Duplicate Finder compares by similarity instead, spotting copies that have been resized, re-encoded, or had a watermark added. The optional AI pass goes further, using neural embeddings to find cropped, mirrored, or zoomed versions that pixel comparison misses, and it can even tell you when a short clip is just a piece of a longer recording. Everything runs locally, and the Docker image plus web UI make it practical on a NAS or headless server. If you manage a large media library and want to reclaim disk space without deleting the wrong file, this is a solid choice.