# facebookarchive/C3D

C3D is a modified version of BVLC caffe to support 3D ConvNets.

Repository: https://github.com/facebookarchive/C3D
Canonical: https://ross.abutalabs.com/products/c3d
Language: Jupyter Notebook
License: NOASSERTION
License Family: other
Archived: true
Last push: 2019-07-31T22:24:07+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 4163, "days_push": 2590, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, archived, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1181, forks 498 (observed 2026-08-28T04:03:54.078027+00:00)

## What it is
C3D is a modified fork of BVLC Caffe that adds support for 3D convolution and pooling layers, enabling training and fine-tuning of 3D ConvNets on video. It also provides pre-trained C3D models for extracting video features.

## Use cases
- train 3D convnets for video classification
- extract video features with pre-trained C3D models
- fine-tune 3D CNNs for action recognition
- spatiotemporal feature learning from videos

## When to choose
- you specifically need the original C3D architecture or its pre-trained models in a Caffe-based pipeline
- reproducing older research that depends on C3D

## When to avoid
- starting a new video understanding project - use the maintained R2Plus1D caffe2 models instead
- you need active support, modern GPU compatibility, or PyTorch/TensorFlow integration
- commercial use - the license is Creative Commons Attribution-NonCommercial 3.0

## Facets
- artifact type: library
- maturity: abandoned
- function: machine-learning, deep-learning, video-processing, image-processing
- domain: machine-learning, deep-learning, computer-vision
- platform: cpp, python
- tags: 3d-convnets, caffe, video-feature-extraction, action-recognition, archived, video, linux, gpu

## Member repositories
- facebookarchive/C3D (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:54.078027+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 2026-08-30T06:25:10.692569+00:00, confidence not recorded.
  - readme: https://github.com/facebookarchive/C3D (fetched 2026-08-28T04:03:54.078027+00:00, sha d5456d7aedc5)
- Data as of 2026-08-30T08:39:29.467469+00:00.