aiqbench analysis
This is an ML method paper: a reliability-weighted Gaussian Product-of-Experts for combining multimodal inputs when some modalities are missing or corrupted, evaluated on two clinical tasks. The reported gains are clearest when modalities are absent, and the authors note failure under severe degradation and that learning reliable quality signals remains open. Readers should treat the results as benchmark-specific evidence on clinical datasets, not a general claim about robustness.
Reliability-Aware Probabilistic Fusion for Multimodal Learning with Incomplete Evidence
arXiv:2610.29621v1 Announce Type: new Abstract: Multimodal inputs may be missing or degraded, raising distinct questions of evidence availability and trust. We introduce a reliability-aware Gaussian Product-of-Experts (PoE) in which an input-dependent reliability weight tempers each modality's latent precision, while missing experts are omitted. We characterize how reliability and availability affect posterior information and expert influence. Across two clinical multimodal tasks, the full framework remains competitive with complete inputs and achieves its clearest mean-AUROC advantage under modality absence, particularly on HAIM-MIMIC-MM. Ablations show a substantial gain from missing-aware training and an additional HAIM improvement from reliability weighting under matched PoE, missing-aware, and consistency training. Unseen test-time corruptions are used diagnostically and reveal modality-dependent reliability-weight and latent-uncertainty responses, including failure under severe degradation. Thus, reliability weighting provides an explicit evidence-control mechanism, while learning dependable quality signals remains an open challenge.
Abstract:Multimodal inputs may be missing or degraded, raising distinct questions of evidence availability and trust. We introduce a reliability-aware Gaussian Product-of-Experts (PoE) in which an input-dependent reliability weight tempers each modality's latent precision, while missing experts are omitted. We characterize how reliability and availability affect posterior information and expert influence. Across two clinical multimodal tasks, the full framework remains competitive with complete inputs and achieves its clearest mean-AUROC advantage under modality absence, particularly on HAIM-MIMIC-MM. Ablations show a substantial gain from missing-aware training and an additional HAIM improvement from reliability weighting under matched PoE, missing-aware, and consistency training. Unseen test-time corruptions are used diagnostically and reveal modality-dependent reliability-weight and latent-uncertainty responses, including failure under severe degradation. Thus, reliability weighting provides an explicit evidence-control mechanism, while learning dependable quality signals remains an open challenge.
Machine Learning (cs.LG)
arXiv:2610.29621 [cs.LG]
(or arXiv:2610.29621v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2610.29621 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 227 words