# piergiaj/pytorch-i3d

Repository: https://github.com/piergiaj/pytorch-i3d
Canonical: https://ross.abutalabs.com/products/pytorch-i3d
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2020-06-28T01:50:12+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": 3082, "days_push": 2258, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1054, forks 260 (observed 2026-08-28T04:03:24.103899+00:00)

## What it is
A PyTorch port of DeepMind's I3D (Inflated 3D ConvNet) models pretrained on the Kinetics dataset for video action recognition. It includes code for fine-tuning on Charades and extracting per-segment video features as numpy arrays.

## Use cases
- recognize human actions in videos
- extract I3D features from video clips
- fine-tune a video action recognition model on my own dataset
- convert DeepMind Kinetics I3D weights to PyTorch
- classify video segments on Charades
- get per-segment embeddings for video understanding tasks

## When to choose
- you need a classic pretrained video action recognition backbone in PyTorch
- you want to extract I3D features for downstream models
- you are reproducing the Quo Vadis action recognition paper or Charades challenge results

## When to avoid
- you need support for recent PyTorch versions (code targets PyTorch 0.3)
- you want state-of-the-art video models rather than a 2017-era architecture
- you need maintained software with active bug fixes

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, video-processing, computer-vision
- domain: deep-learning, computer-vision
- platform: python
- tags: i3d, action-recognition, kinetics, pytorch, pretrained-models, feature-extraction, video-understanding, video

## Member repositories
- piergiaj/pytorch-i3d (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:24.103899+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 2026-08-30T06:59:02.484821+00:00, confidence not recorded.
  - readme: https://github.com/piergiaj/pytorch-i3d (fetched 2026-08-28T04:03:24.103899+00:00, sha f69475ac1943)
- Data as of 2026-08-30T08:39:29.467469+00:00.