uva-cv-lab/OmniShotCut_v1.5

OmniShotCut: Holistic Relational Shot Boundary Detection with Shot-Query Transformer

Authors: Boyang Wang, Guangyi Xu, Jiahui Zhang, Zhipeng Tang, Zezhou Cheng
Affiliation: UVA Computer Vision Lab

OmniShotCut is a sensitive and informative shot boundary detector. It detects shot changes across diverse sources (anime, vlog, game, shorts, sports, screen recording, etc.) and recognizes Sudden Jumps and Transitions (dissolve, fade, wipe, ...) with a Shot-Query-based Video Transformer.

Model Zoo

Model	Description	Link
OmniShotCut v1.5 (recommended)	Hidden dim 576, 52.8M params. Best real-world quality: fewer false cuts on continuous camera motion / lighting changes, and better dense fast-cut recall.	Download
OmniShotCut (paper)	3 encoders, 34.5M params. Reproduces the paper (Table 1) numbers.	Download
OmniShotCut (preview)	6 encoders, 41M params. Early preview release.	Download

Usage

Install:

pip install git+https://github.com/UVA-Computer-Vision-Lab/OmniShotCut.git

Run shot boundary detection:

import omnishotcut

cut_model = omnishotcut.load("uva-cv-lab/OmniShotCut_v1.5", filename="OmniShotCut_ckpt.pth")
ranges = cut_model.inference("video.mp4", mode="clean_shot")

Citation

@article{wang2026omnishotcut,
  title={OmniShotCut: Holistic Relational Shot Boundary Detection with Shot-Query Transformer},
  author={Wang, Boyang and Xu, Guangyi and Zhang, Jiahui and Tang, Zhipeng and Cheng, Zezhou},
  journal={arXiv preprint arXiv:2604.24762},
  year={2026}
}