# s9xie/hed

code for Holistically-Nested Edge Detection

Repository: https://github.com/s9xie/hed
Canonical: https://ross.abutalabs.com/products/hed
Language: C++
License: NOASSERTION
License Family: other
Last push: 2024-04-07T16:13:16+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 3989, "days_push": 878, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1925, forks 534 (observed 2026-08-28T04:05:55.534642+00:00)

## What it is
Reference implementation of Holistically-Nested Edge Detection (HED), a deep learning edge and boundary detection model built on a modified Caffe framework with fully convolutional networks. It includes pretrained models, training, and testing code for reproducing benchmark results on BSDS500 and NYU Depth.

## Use cases
- detect edges in images with a deep learning model
- reproduce HED results on the BSDS500 benchmark
- train an edge detection network on custom data
- extract object boundaries from images
- compare edge detection algorithms against HED baselines

## When to choose
- you need state-of-the-art learned edge/boundary detection rather than classical operators like Canny
- you want pretrained weights and reproducible benchmark results
- your stack already uses Caffe

## When to avoid
- you need a maintained framework on modern PyTorch/TensorFlow stacks
- you want a plug-and-play production library rather than research code
- you cannot build legacy Caffe dependencies

## Facets
- artifact type: library
- maturity: maintenance
- function: computer-vision, image-processing, deep-learning
- domain: computer-vision, image-processing, deep-learning
- platform: cpp, python
- tags: edge-detection, caffe, fully-convolutional-networks, research-code, boundary-detection, linux

## Member repositories
- s9xie/hed (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:55.534642+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 2026-08-30T03:09:22.499049+00:00, confidence not recorded.
  - readme: https://github.com/s9xie/hed (fetched 2026-08-28T04:05:55.534642+00:00, sha d36a73eefd14)
- Data as of 2026-08-30T08:39:29.467469+00:00.