mpatacchiola/deepgaze
A Python computer vision library implementing CNN-based head pose estimation, gaze direction estimation, face detection, motion tracking, and saliency mapping for human-computer interaction.
Not currently ranked — collecting fresh signals.
star history
Deepgaze is a library for human-computer interaction that uses Convolutional Neural Networks to detect faces, estimate head pose and gaze direction, track motion, and generate saliency maps. The focus is on approximating attention by analyzing head orientation, which is useful when eyes are not visible or when the user is far from the camera. It also includes implementations for skin detection through backprojection and motion detection using particle filters.
Frequently asked
- What is mpatacchiola/deepgaze?
- A Python computer vision library implementing CNN-based head pose estimation, gaze direction estimation, face detection, motion tracking, and saliency mapping for human-computer interaction.
- Is deepgaze open source?
- Yes — mpatacchiola/deepgaze is open source, released under the MIT license.
- What language is deepgaze written in?
- mpatacchiola/deepgaze is primarily written in Python.
- How popular is deepgaze?
- mpatacchiola/deepgaze has 1.9k stars on GitHub.
- Where can I find deepgaze?
- mpatacchiola/deepgaze is on GitHub at https://github.com/mpatacchiola/deepgaze.