Ben-Gurion University of the Negev (Ifergane, Amoyal, Benishay, Freifeld) · Multi-camera video stitching

MViS: Video Stitching from Multiple Moving Cameras

MViS (Multiple Video Stitching) is the official implementation of a NeurIPS 2026 paper from Ben-Gurion University of the Negev. It is a test-time optimisation method that aligns videos from any number of freely moving cameras, with no known camera ordering and a view-intersection graph that may change over time, and produces one panoramic video. The repository includes a demo script, a training and evaluation script that runs on the MMCD and StabStitch-D datasets, and the GV metric. It assumes the videos are roughly synchronised and that every video overlaps with at least one other. The page gives no quantitative results and no licence.

Open the source ↗

Similar Video & Animation releases

  • Reshoot Anything

    Adithya Iyer et al. · Video Editing · Week 24, 2026

  • SCAIL-2

    Z.ai (zai-org) · Video Editing · Week 24, 2026

  • EffectMaker

    Research · Video Editing · Week 11, 2026

  • VINO

    Research · Text-to-Video · Week 2, 2026

  • UniVideo

    KwaiVGI · Text-to-Video · Week 2, 2026

  • SteadyDancer

    NJU MCG · Video Editing · Week 50, 2025

Published by Quanta AI Labs, an AI agency in India. Browse every week in AI News Weekly.