# CSAILVision/GazeCapture

Eye Tracking for Everyone

Repository: https://github.com/CSAILVision/GazeCapture
Canonical: https://ross.abutalabs.com/products/gazecapture
Homepage: http://gazecapture.csail.mit.edu
Language: Python
License: NOASSERTION
License Family: other
Last push: 2023-07-06T21:43:46+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3510, "days_push": 1154, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1069, forks 263 (observed 2026-08-28T04:03:27.736963+00:00)

## What it is
GazeCapture is a large-scale eye tracking dataset with ~2.5M frames from over 1450 subjects, released alongside the iTracker CNN model and Caffe code from the CVPR 2016 paper 'Eye Tracking for Everyone'. It enables gaze prediction on commodity mobile devices without extra sensors.

## Use cases
- train a gaze estimation model on mobile phone images
- download a large eye tracking dataset for research
- reproduce the iTracker CNN from the CVPR 2016 paper
- benchmark gaze prediction error on phones and tablets
- study face and eye bounding box metadata for gaze datasets

## When to choose
- you need large-scale real-world gaze data for training or benchmarking
- you want a research-licensed dataset with pretrained Caffe models
- you are reproducing or comparing against iTracker results

## When to avoid
- you need a permissively licensed dataset for commercial use
- you need actively maintained PyTorch or TensorFlow code
- you need real-time eye tracking out of the box without training

## Facets
- artifact type: dataset
- maturity: maintenance
- function: computer-vision, machine-learning, deep-learning
- domain: computer-vision, machine-learning, artificial-intelligence
- platform: python, cross-platform
- tags: eye-tracking, gaze-estimation, dataset, caffe, itracker, research

## Member repositories
- CSAILVision/GazeCapture (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:27.736963+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:54:15.404592+00:00, confidence not recorded.
  - readme: https://github.com/CSAILVision/GazeCapture (fetched 2026-08-28T04:03:27.736963+00:00, sha 7c1a27b2e147)
  - homepage: http://gazecapture.csail.mit.edu (fetched 2026-08-29T12:56:46.813827+00:00, sha 8b80efdfdde6)
- Data as of 2026-08-30T08:39:29.467469+00:00.