ImageAI

ImageAI is the Python library for rapid, self-contained Deep Learning and Computer Vision implementation: execute powerful tasks in just a few lines of code.

ImageAI provides direct access to state-of-the-art Computer Vision, built on a robust PyTorch backend. The library empowers developers to implement core functions like Image Recognition (1000+ objects) and Object Detection (80 COCO objects) using pre-trained models such as RetinaNet and YOLOv3. It handles video analysis, object tracking, and custom model training with minimal code, making complex Deep Learning accessible for immediate application development.

https://imageai.readthedocs.io/en/latest/