Skip to content

pawkor/ai-autoedit

Repository files navigation

AI-autoedit — AI Highlight Reel Pipeline

Każdy dzień na motocyklu to kilkaset gigabajtów surowego materiału z jednej lub dwóch kamer. Ręczny montaż zajmuje wielokrotnie więcej czasu niż sam wyjazd.

AI-autoedit rozwiązuje ten problem: model CLIP ocenia każdą scenę semantycznie — rozumie co jest w kadrze, nie tylko ruch czy jasność — i automatycznie składa highlight reel z muzyką. Próg selekcji, manualne overrides klatek i dobór muzyki dostępne są przez przeglądarkę bez dotykania terminala.


Every day on a motorcycle produces hundreds of gigabytes of raw footage from one or two cameras. Manual editing takes longer than the ride itself.

AI-autoedit solves this: a CLIP model scores each scene semantically — understanding what's in the frame, not just motion or brightness — and automatically assembles a highlight reel with music. Selection threshold, manual frame overrides, and music selection are all accessible through the browser without touching the terminal.


Przykładowy film / Sample output

Przykładowy highlight

Dokumentacja / Documentation

Web UI — kolejność pracy / workflow order

Techniczne / Technical


Licencja / License

AGPL-3.0 + Commons Clause — możesz używać, modyfikować i hostować lokalnie za darmo. Komercyjne użycie i SaaS są zabronione. → Licencja / License

Prywatność / Privacy

Działa wyłącznie na Twoim sprzęcie. Brak telemetrii i wymogu konta. → Prywatność / Privacy

Zastrzeżenie / Disclaimer

PL: Zrób kopię zapasową materiału przed użyciem. Autor nie ponosi odpowiedzialności za utratę danych ani uszkodzenie plików. Pełna klauzula: LICENSE.

EN: Back up your footage before use. The author is not responsible for data loss or file corruption. See LICENSE.

About

AI pipeline that turns a day's motorcycle footage into a highlight reel — no manual editing. CLIP model scores every scene semantically, selects the best shots, interleaves dual-camera footage and picks music by energy level.

Topics

Resources

License

Stars

Watchers

Forks

Contributors