AI summaryⓘ
The authors introduce a new method called C²MF that helps combine information from different sources more smartly by checking how trustworthy each source is depending on the situation. Unlike older methods that assume each source is always equally reliable, their approach uses a special model, Conditional Probabilistic Circuit, to measure reliability on a case-by-case basis. They also create a new way to calculate reliability called Context-Specific Information Credibility, which adapts to changing conditions. They test their method on a new benchmark designed to create conflicts between sources and show it can improve accuracy by a lot when there’s noise or corruption. Their method also keeps things understandable by using probabilistic circuits.
Multimodal FusionConditional Probabilistic CircuitContext-Specific Information CredibilityKL-DivergenceSource ReliabilityCross-Modal ConflictsClass-Specific CorruptionRobustnessProbabilistic Circuits
Authors
Pranuthi Tenali, Sahil Sidheekh, Saurabh Mathur, Erik Blasch, Kristian Kersting, Sriraam Natarajan
Abstract
Multimodal fusion requires integrating information from multiple sources that may conflict depending on context. Existing fusion approaches typically rely on static assumptions about source reliability, limiting their ability to resolve conflicts when a modality becomes unreliable due to situational factors such as sensor degradation or class-specific corruption. We introduce C$^2$MF, a context-specfic credibility-aware multimodal fusion framework that models per-instance source reliability using a Conditional Probabilistic Circuit (CPC). We formalize instance-level reliability through Context-Specific Information Credibility (CSIC), a KL-divergence-based measure computed exactly from the CPC. CSIC generalizes conventional static credibility estimates as a special case, enabling principled and adaptive reliability assessment. To evaluate robustness under cross-modal conflicts, we propose the Conflict benchmark, in which class-specific corruptions deliberately induce discrepancies between different modalities. Experimental results show that C$^2$MF improves predictive accuracy by up to 29% over static-reliability baselines in high-noise settings, while preserving the interpretability advantages of probabilistic circuit-based fusion.