fizyr/keras-retinanet
A Keras/TensorFlow implementation of RetinaNet for object detection, training on bounding box predictions with focal loss.
Not currently ranked — collecting fresh signals.
star history
This repository provides a Keras implementation of the RetinaNet object detection architecture as described in the original paper by Lin et al. It implements the focal loss mechanism to address class imbalance in dense object detection and uses a Feature Pyramid Network (FPN) backbone for multi-scale feature extraction. The model accepts image inputs and outputs bounding box coordinates, confidence scores, and class labels for detected objects.
Frequently asked
- What is fizyr/keras-retinanet?
- A Keras/TensorFlow implementation of RetinaNet for object detection, training on bounding box predictions with focal loss.
- Is keras-retinanet open source?
- Yes — fizyr/keras-retinanet is open source, released under the Apache-2.0 license.
- What language is keras-retinanet written in?
- fizyr/keras-retinanet is primarily written in Python.
- How popular is keras-retinanet?
- fizyr/keras-retinanet has 4.4k stars on GitHub.
- Where can I find keras-retinanet?
- fizyr/keras-retinanet is on GitHub at https://github.com/fizyr/keras-retinanet.