CatSignal models context as a Bayesian prior in a Product-of-Experts setup to infer cat intents from multimodal observations, reaching 77.72% accuracy while reducing context shortcut errors.
Baltru ˇsaitis, C
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Context as Prior: Bayesian-Inspired Intent Inference for Non-Speaking Agents with a Household Cat Testbed
CatSignal models context as a Bayesian prior in a Product-of-Experts setup to infer cat intents from multimodal observations, reaching 77.72% accuracy while reducing context shortcut errors.