Hi everyone,
I’m working with MediaPipe’s gesture detector and I’m looking for advice on finding different gesture models.
Here are the gestures I’m trying to detect:
Sounds like a job for Wekinator@visionnocturne . With Wekinator you basically make your own model. So you could, for example, take the angles between the finger joints and use those as ‘features’ for the ‘labels’ you’ve described above-