README.md

metadata

license: apache-2.0
library_name: pytorch
pipeline_tag: video-classification
pretty_name: SurgMotion
gated: true
extra_gated_heading: SurgMotion Access Agreement
extra_gated_description: >-
  Access to SurgMotion is provided under the same license as the GitHub
  repository: Apache License 2.0. Please review the license at
  https://github.com/CAIR-HKISI/SurgMotion/blob/main/LICENSE and provide your
  institution information before requesting access.
extra_gated_button_content: Agree and request access
extra_gated_fields:
  Institution / Organization: text
  Department / Lab: text
  Position / Role:
    type: select
    options:
      - Student
      - Researcher
      - Faculty
      - Clinician
      - Engineer
      - Other
  Country / Region: country
  Intended use:
    type: select
    options:
      - Academic research
      - Education
      - Clinical research
      - Benchmarking
      - Other
  Project description: text
  I have read and agree to the Apache License Version 2 terms for SurgMotion: checkbox
tags:
  - surgical-video
  - foundation-model
  - video-understanding
  - phase-recognition
  - v-jepa
  - vision-transformer
  - pytorch
datasets:
  - SurgMotion-15M
language:
  - en

SurgMotion-G

SurgMotion-G is the ViT-Giant-xformer checkpoint of SurgMotion, a video-native foundation model for universal understanding of surgical videos.

This repository contains a single released checkpoint so that the model is structured as a standard single-model Hugging Face repository.

Files

• config.json: minimal repository config for Hub compatibility
• SurgMotion-vitg-xformer.safetensors: recommended weights in safetensors format
• SurgMotion-vitg-xformer.pt: original released PyTorch checkpoint retained for compatibility

Download

Download the repository config and safetensors checkpoint with huggingface_hub:

from huggingface_hub import hf_hub_download

config_path = hf_hub_download(
    repo_id="CAIR-HKISI/SurgMotion-vitg-xformer",
    filename="config.json",
)

weights_path = hf_hub_download(
    repo_id="CAIR-HKISI/SurgMotion-vitg-xformer",
    filename="SurgMotion-vitg-xformer.safetensors",
)

If your existing code expects the original PyTorch checkpoint format, keep using SurgMotion-vitg-xformer.pt.

Project Links

• Project page: https://surgmotion.cares-copilot.com/
• Paper: https://arxiv.org/abs/2602.05638
• Code: https://github.com/CAIR-HKISI/SurgMotion
• Full model collection: https://huggingface.co/CAIR-HKISI/SurgMotion