Recognize Hand Gesture

How it fits

Typical backends

Caveats

• I/O contractEvery frame emits exactly one output. The gesture vocabulary is fixed by the bound bundle; on the list arm an empty list, and on the single arm a zero-confidence empty label, both mean no hand was found. A frame that cannot be read at all degrades to that same empty result rather than being dropped.
• LimitationOne gesture is named per detected hand, not a ranked set of alternatives for one hand - the recognizer's classification head emits a single winning category per hand. The list therefore holds at most num_hands entries, ordered by confidence.
• Parameter interactionnum_hands caps the hand detector, which picks which hands to keep by how confidently it locates them, not by how confidently their gesture is named. Lowering it therefore drops a hand the detector ranked lower even when that hand carried the higher gesture confidence, so the leading entry of the list can change with it. Set it to the most hands a frame should ever carry rather than tuning it down to shorten the list.
• AccuracyThe eight recognised gestures are a closed set. Anything outside it is named None, which is a real classification rather than an error - read None as "a hand is present but holds no recognised gesture".
• LatencyInference runs on CPU with the XNNPACK backend. The task layer is not re-entrant, so concurrent frames are serialized through one recognizer; a frame costs single-digit to low-tens of milliseconds at typical camera resolutions.
• State lifetimeThe bundle bound at model is opened once at startup and the recognizer lives for the process. num_hands and the three confidence floors are fixed into the graph at that point and cannot change without a redeploy; only score_threshold is re-read per frame.