SciVid STIR: Surgical Tattoos in Infrared Point-Tracking Clips of Soft Tissue in Endoscopic Video (Google DeepMind)
- Provider
- Google DeepMind (SciVid); source data University of British Columbia and Intuitive Surgical
- License
- CC-BY-4.0 (source)
- Stored at
- Google Cloud Storage, gs://scivid/full/stir (us)
- Access
- Free and public. No account with the data provider is needed.
- Category
- Health & Biomedical
- Source
- https://arxiv.org/abs/2309.16782
About this dataset
Preprocessed STIR (Surgical Tattoos in Infrared) clips for point tracking on deforming soft tissue in surgical endoscopy video, as packaged in the SciVid benchmark. STIR, from the University of British Columbia and Intuitive Surgical, marks tissue with infrared-visible tattoo points that are invisible in the visible-light video, giving ground-truth start and end positions of tracked points across surgical sequences. SciVid filtered and downsampled the clips and normalized query and target points by image size. The folder holds three sets of 100 shards: full_val_stir, short_val_stir and test_stir. There is no training split; models are evaluated on how well tracking transfers to tissue motion in minimally invasive surgery. SciVid (Hasson et al., SciVid: Cross-Domain Evaluation of Video Models in Scientific Applications, ICCV 2025, arXiv:2507.03578) is a five-task benchmark for video foundation models in science; this folder is one task. Files are ArrayRecord shards, read with the array_record Python package or the loaders in the google-deepmind/scivid repository. License: SciVid materials CC BY 4.0 (Google DeepMind); STIR is under a Creative Commons Attribution license at IEEE DataPort (DOI 10.21227/w8g4-g548). Cite SciVid and the STIR paper, arXiv:2309.16782.
How to access this dataset in Python
import gcsfs
fs = gcsfs.GCSFileSystem(token='anon')
fs.ls('scivid/full/stir/', detail=False)[:10]Lists the files with anonymous access. Runs on any machine, with no cloud account.
Data structure
Files
Example file: full/stir/
Analyze this dataset without downloading it
On Scigantic, SciVid STIR: Surgical Tattoos in Infrared Point-Tracking Clips of Soft Tissue in Endoscopic Video (Google DeepMind) opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.
This page is also available as Markdown and as JSON.
Related Health & Biomedical datasets
- NuCLS: Nucleus Classification, Localization and Segmentation in TCGA breast cancer H&E (corrected single-rater set, 59k nuclei)
- IDC TCGA-COAD: Colon adenocarcinoma whole-slide microscopy, CT, other (464 patients)
- IDC PROSTATE-MRI: Prostate Cancer MRI (26 patients)
- HVSMR-2.0: 3D cardiovascular MRI with whole-heart segmentations in congenital heart disease (60 patients)
- IDC ISPY2: Breast Cancer MRI (719 patients)
- Salmonella Serovar Hyperspectral Microscopy: 500 Hypercube Images of Five Serovars for AI Classification
- IDC CMB-PCA: Prostate Cancer CT, MRI, whole-slide microscopy, nuclear medicine (54 patients)
- IDC CPTAC-BRCA: Breast Cancer whole-slide microscopy (198 patients)