Multi-modal fusion — camera + EEG/EMG combined classification #22
Labels
No labels
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference: Circuit-Forge/raven#22
Loading…
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Combine camera-based gesture confidence and EEG/EMG signal confidence into a single fused decision.
Architecture: weighted voting across active sources. Each source produces a
(class, confidence)pair per frame. Fuser computes weighted sum; fires action only when aggregate confidence > threshold.Weights: tunable per profile. Default: camera 0.6, EEG 0.3, EMG 0.1 (camera is most reliable baseline). User can adjust in config.
Fusion modes:
any— fire when any single source reaches threshold (low latency, higher false-positive rate)majority— fire when weighted majority agrees (default, balanced)unanimous— all active sources must agree (lowest false-positive, higher latency)Benefit: camera confirms EEG intent for ambiguous epochs. EEG confirms camera gesture for low-light or partial occlusion. Combined accuracy significantly higher than either source alone.
This is the Phase C flagship feature — the full Merlin stack operating simultaneously.