VIAME/VIAME

Video and Image Analytics for Multiple Environments

What it solves

VIAME is a comprehensive computer vision toolkit designed for "do-it-yourself" AI. It provides a unified platform for users to perform object detection, tracking, image/video annotation, and size measurement without needing to be an expert in every underlying algorithm. It was originally created for marine species analytics but now serves as a general-purpose vision toolkit.

How it works

VIAME uses a multi-threaded pipeline framework (via kwiver) that can connect nodes written in C++, Python, and Matlab. This allows it to integrate a wide variety of algorithms and plugins—including those from PyTorch, TensorFlow, TensorRT, and Darknet (YOLO)—into a single workflow. It offers both desktop and web-based user interfaces (DIVE, VIEW, and SEARCH) for annotation, training, and video search.

Who it’s for

It is designed for researchers and developers who need to process visual data, annotate images/videos, and rapidly generate models for object detection and tracking, particularly those working in specialized domains like marine biology or general computer vision tasks.

Highlights

  • Multi-Framework Support: Integrates detectors and trackers from PyTorch, TensorFlow, TensorRT, and Darknet.
  • Extensible Pipeline: A multi-threaded framework that allows the connection of C++, Python, and Matlab nodes.
  • End-to-End Workflow: Includes tools for image enhancement, mosaicing, stereo measurement, and algorithm evaluation.
  • Flexible Deployment: Available as a full desktop installation, a web-based annotator, or headless Docker images for GPU-accelerated processing.

Related

  • Project
    wkentaro/labelme

    labelme is a cross‑platform Python/Qt desktop app for drawing bounding boxes, polygons, masks, etc., on images or video. Annotations are saved as JSON and can be exported to VOC or COCO formats, making it a practical tool for creating training data for computer‑vision models. Install via pip, a paid standalone executable, or Linux packages; recent versions add AI‑assisted mask generation (SAM, YOLO‑World).

  • Project
    ok-oldking/ok-script

    A computer vision-based Python automation framework for Windows apps, games, and Android devices that supports OCR and template matching.

  • Project
    cvat-ai/cvat

    An open-source data annotation platform for building high-quality visual datasets for computer vision, supporting image, video, and 3D annotation.

  • Project
    oculix-org/SikuliX1

    A computer vision-based automation tool that identifies and interacts with screen elements using image recognition to simulate mouse and keyboard input.

  • Project
    roboflow/supervision

    A computer vision toolkit that provides model-agnostic building blocks for data loading, visualization, and dataset management to accelerate the development of vision applications.