Controlled Acquisition and Abstention in Three-Channel Score Conflicts
Matched controls show that a threshold policy matches always-request decisions with fewer requests; its advantage over always-answer fusion depends on the reward assigned to that ambiguity.
ProofPaper ↗
Key points
- When audio, video, and text disagree, accuracy alone does not show whether to acquire another source or abstain.
- We study these choices in a controlled three-score benchmark: a policy observes two signed scores, may request the third at a cost, and can abstain.
- On held-out-actor emotion clips, eight-frame fusion has opposite-signed accuracy differences for two encoder pairs, with both actor intervals containing zero; matched-request routing gains are small and uncertain.
- These results separate full-modality accuracy from pre-request selection value and show that selection value depends on budget and the observed-pair ranking.
Sources (1)
- [1]Controlled Acquisition and Abstention in Three-Channel Score ConflictsarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 7, 07:09 PM
Matched controls show that a threshold policy matches always-request decisions with fewer requests; its advantage over always-answer fusion depends on the reward assigned to that ambiguity.
When audio, video, and text disagree, accuracy alone does not show whether to acquire another source or abstain.
Extractive summary: sentences quoted from the sources.