willem.com

Controlling video with human pose recognition

Using TensorFlow and PoseNet on a video feed

Dec. 1, 2019 -

For an exciting new project I have been experimenting with computer vision using TensorFlow. I wanted to achieve realtime human pose detection to drive interactive video projections and games. Time to dive into the world of machine learning, tensors and computer vision!

Continue reading
MobileNet-SSD algorithm (image: hey-yahei.cn)
MobileNet-SSD algorithm (image: hey-yahei.cn)
ReCaptcha: you've been training Google's AI for years!
ReCaptcha: you've been training Google's AI for years!
Me and mini-me being recognised by the PoseNet model
Me and mini-me being recognised by the PoseNet model
PomPom faux puff mirror by Daniel Rozin
PomPom faux puff mirror by Daniel Rozin
Turning people into birds using computer vision
Turning people into birds using computer vision

Other stories

all blog posts

Tagged

DataExperimentGamePeopleSensorsWork