Start with the framework overview, then follow full tutorials for building your own gesture recognizers — from capturing a dataset to running live detection.
The big picture: how the free desktop capture app and the Python training package fit together to build gesture, pose, and activity recognition — entirely on your machine.
Demo
Real-Time ASL Fingerspelling Recognition (A–Z)
Gesto recognising the full American Sign Language alphabet, live — a real end-to-end example of what you can build.
Labeller
How to Use Gesto Labeller — Full Walkthrough
A complete tour of the desktop app: projects, capturing samples, exploring and editing your data, and exporting it ready to train.
Tutorial · Static
Build a Static Gesture Recognizer — Capture, Train & Detect
End to end: capture a static (held-shape) dataset, train the model, and run live detection. Perfect for hand signs and postures.
Tutorial · Sequence
Build a Sequence Gesture Recognizer — Detect Motion & Actions
End to end for motion gestures: capture clips, train an LSTM model, and detect dynamic actions like waving or walking.
Labeller · Data
Capture a Static Dataset + Import & Export
Build a static dataset from your webcam, then save, back up, and move it with the import and export features.
Labeller · Data
Capture a Sequence Gesture Dataset in Gesto Labeller
Record motion clips with consistent length, ready for training a sequence model.