Credibility-Aware Multimodal Fusion Using Probabilistic Circuits

Sahil Sidheekh, Pranuthi Tenali, Saurabh Mathur, Erik Blasch, Kristian Kersting, Sriraam Natarajan
Proceedings of The 28th International Conference on Artificial Intelligence and Statistics, PMLR 258:2305-2313, 2025.

Abstract

We consider the problem of late multimodal fusion for discriminative learning. Motivated by noisy, multi-source domains that require understanding the reliability of each data source, we explore the notion of credibility in the context of multimodal fusion. We propose a combination function that uses probabilistic circuits (PCs) to combine predictive distributions over individual modalities. We also define a probabilistic measure to evaluate the credibility of each modality via inference queries over the PC. Our experimental evaluation demonstrates that our fusion method can reliably infer credibility while being competitive with the state-of-the-art.

Cite this Paper

BibTeX
@InProceedings{pmlr-v258-sidheekh25a, title = {Credibility-Aware Multimodal Fusion Using Probabilistic Circuits}, author = {Sidheekh, Sahil and Tenali, Pranuthi and Mathur, Saurabh and Blasch, Erik and Kersting, Kristian and Natarajan, Sriraam}, booktitle = {Proceedings of The 28th International Conference on Artificial Intelligence and Statistics}, pages = {2305--2313}, year = {2025}, editor = {Li, Yingzhen and Mandt, Stephan and Agrawal, Shipra and Khan, Emtiyaz}, volume = {258}, series = {Proceedings of Machine Learning Research}, month = {03--05 May}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v258/main/assets/sidheekh25a/sidheekh25a.pdf}, url = {https://proceedings.mlr.press/v258/sidheekh25a.html}, abstract = {We consider the problem of late multimodal fusion for discriminative learning. Motivated by noisy, multi-source domains that require understanding the reliability of each data source, we explore the notion of credibility in the context of multimodal fusion. We propose a combination function that uses probabilistic circuits (PCs) to combine predictive distributions over individual modalities. We also define a probabilistic measure to evaluate the credibility of each modality via inference queries over the PC. Our experimental evaluation demonstrates that our fusion method can reliably infer credibility while being competitive with the state-of-the-art.} }
Endnote
%0 Conference Paper %T Credibility-Aware Multimodal Fusion Using Probabilistic Circuits %A Sahil Sidheekh %A Pranuthi Tenali %A Saurabh Mathur %A Erik Blasch %A Kristian Kersting %A Sriraam Natarajan %B Proceedings of The 28th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2025 %E Yingzhen Li %E Stephan Mandt %E Shipra Agrawal %E Emtiyaz Khan %F pmlr-v258-sidheekh25a %I PMLR %P 2305--2313 %U https://proceedings.mlr.press/v258/sidheekh25a.html %V 258 %X We consider the problem of late multimodal fusion for discriminative learning. Motivated by noisy, multi-source domains that require understanding the reliability of each data source, we explore the notion of credibility in the context of multimodal fusion. We propose a combination function that uses probabilistic circuits (PCs) to combine predictive distributions over individual modalities. We also define a probabilistic measure to evaluate the credibility of each modality via inference queries over the PC. Our experimental evaluation demonstrates that our fusion method can reliably infer credibility while being competitive with the state-of-the-art.
APA
Sidheekh, S., Tenali, P., Mathur, S., Blasch, E., Kersting, K. & Natarajan, S.. (2025). Credibility-Aware Multimodal Fusion Using Probabilistic Circuits. Proceedings of The 28th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 258:2305-2313 Available from https://proceedings.mlr.press/v258/sidheekh25a.html.

Related Material