3 ms·TriCycle: Audio Representation Learning from Sensor Data via Self-Supervision2 points by jarmitage 7y ago