Unified Multimodal Uncertain Inference
This work addresses the lack of a framework for fine-grained probabilistic reasoning across multiple modalities, enabling calibrated uncertainty estimation in multimodal AI.
The paper introduces Unified Multimodal Uncertain Inference (UMUI), a task requiring calibrated probability estimates across text, audio, and video. The proposed CLUE method achieves performance comparable to or better than baselines up to 32B parameters with a 3B model.
We introduce Unified Multimodal Uncertain Inference (UMUI), a multimodal inference task spanning text, audio, and video, where models must produce calibrated probability estimates of hypotheses conditioned on a premise in any modality or combination. While uncertain inference has been explored in text, extension to other modalities has been limited to single-modality binary entailment judgments, leaving no framework for fine-grained probabilistic reasoning in or across other modalities. To address this, we curate a human-annotated evaluation set with scalar probability judgments across audio, visual, and audiovisual settings, and additionally evaluate on existing text and audio benchmarks. We introduce CLUE (Calibrated Latent Uncertainty Estimation), which combines self-consistent teacher calibration and distribution-based confidence probing to produce calibrated predictions. We demonstrate that our 3B-parameter model achieves equivalent or stronger performance than baselines up to 32B parameters across all modalities.