Perception weights
Weights that watch and listen: objects in a video frame, events in audio, motion across a scene. Fed through the compound eye and the antennae of the connectome.
Perception weights use the visual and auditory afferents of the connectome as their input layer. Frames and audio are encoded by a standard adapter into the 19,262 afferent neurons; the answer is read from a small decoder on the intrinsic population.
GazeObject detection · betaGaze detects and tracks objects in video through the connectome's visual system, with a classical CNN co-processor for class labels. Built for motion, so it finds what moves first.SonoAudio events · betaSono detects sound events in a live stream: a door, a bark, a fall, a name. Audio enters through the auditory afferents of the connectome; the answer is an event with a timestamp.TrackMotion tracking · betaTrack follows a chosen object through a video with the connectome's motion detectors: no class labels, no re-detection, just a point and a velocity, every frame.
Pricing
Every weight set is free up to 100k monthly active devices per platform. Unlimited inference per user.