# symisc/sod

An Embedded Computer Vision & Machine Learning Library (CPU Optimized & IoT Capable)

Repository: https://github.com/symisc/sod
Canonical: https://ross.abutalabs.com/products/sod
Homepage: https://sod.pixlab.io
Language: C
License: NOASSERTION
License Family: other
Topics: computer-vision, library, deep-learning, image-processing, object-detection, c, cpu, real-time, convolutional-neural-networks, face-detection, facial-landmarks, machine-learning-algorithms, image-recognition, image-analysis, vision-framework, embedded, detection, iot-device, iot, webassembly
Last push: 2023-10-20T18:32:09+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 3067, "days_push": 1048, "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 1798, forks 210 (observed 2026-08-28T04:05:37.723632+00:00)

## What it is
SOD is an embedded, cross-platform computer vision and machine learning library written in C, distributed as a single dependency-free amalgamated file with pre-trained models. It provides real-time object detection, face detection, and over 100 image processing routines optimized for CPU-only and resource-constrained IoT/embedded devices.

## Use cases
- detect faces and objects in real time on cpu-only devices
- run object detection on embedded systems or iot devices without gpu
- extract license plates from images
- track eyes and pupils or extract facial landmarks
- apply image filters like gaussian blur, sepia, and edge detection in c
- train custom single-class object detection models
- run computer vision in webassembly

## When to choose
- you need dependency-free, single-file C computer vision for embedded or IoT targets
- you want real-time detection on CPU without GPU or heavy frameworks like OpenCV
- you need pre-trained models plus on-device model training in C

## When to avoid
- you need GPU-accelerated training of large deep learning models
- you prefer Python-first ecosystems like PyTorch or TensorFlow
- you need a fully open-source license - SOD is dual-licensed with commercial terms
- you need multithreaded CNN inference, which is reserved for the commercial version

## Facets
- artifact type: library
- maturity: maintenance
- function: computer-vision, image-processing, machine-learning, deep-learning
- domain: computer-vision, machine-learning, image-processing, embedded-systems, iot
- platform: cross-platform, c, cpp, embedded, iot, wasm, windows
- tags: object-detection, face-detection, cnn, real-time, dependency-free, single-file, pre-trained-models, cpu-optimized, dual-licensed, linux, macos

## Member repositories
- symisc/sod (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:37.723632+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 2026-08-30T03:22:26.807129+00:00, confidence not recorded.
  - readme: https://github.com/symisc/sod (fetched 2026-08-28T04:05:37.723632+00:00, sha 46ec3d5ed96e)
  - homepage: https://sod.pixlab.io (fetched 2026-08-29T11:01:14.719243+00:00, sha 3a610ea96c1a)
  - site_page: https://sod.pixlab.io/changelog.html (fetched 2026-08-29T11:01:14.725617+00:00, sha 19fd93895577)
  - site_page: https://sod.pixlab.io/api.html (fetched 2026-08-29T11:01:14.723276+00:00, sha 6065f891dc8c)
- Data as of 2026-08-30T08:39:29.467469+00:00.