DUAL: Dynamic Uncertainty-Aware Learning

TL;DR

DUAL is a comprehensive framework that improves deep learning performance by effectively modeling and managing feature uncertainties in both single-modal and multi-modal scenarios, leading to significant accuracy gains.

Contribution

It introduces three novel techniques for dynamic uncertainty modeling, distribution-aware modulation, and cross-modal uncertainty learning, advancing the robustness of deep models.

Findings

Achieves 7.1% accuracy improvement on CIFAR-10

Demonstrates 4.1% gain on CMU-MOSEI sentiment analysis

Shows 1.4% accuracy boost on MISR

Abstract

Deep learning models frequently encounter feature uncertainty in diverse learning scenarios, significantly impacting their performance and reliability. This challenge is particularly complex in multi-modal scenarios, where models must integrate information from different sources with inherent uncertainties. We propose Dynamic Uncertainty-Aware Learning (DUAL), a unified framework that effectively handles feature uncertainty in both single-modal and multi-modal scenarios. DUAL introduces three key innovations: Dynamic Feature Uncertainty Modeling, which continuously refines uncertainty estimates through joint consideration of feature characteristics and learning dynamics; Adaptive Distribution-Aware Modulation, which maintains balanced feature distributions through dynamic sample influence adjustment; and Uncertainty-aware Cross-Modal Relationship Learning, which explicitly models uncertainties in cross-modal interactions. Through extensive experiments, we demonstrate DUAL's effectiveness across multiple domains: in computer vision tasks, it achieves substantial improvements of 7.1% accuracy on CIFAR-10, 6.5% accuracy on CIFAR-100, and 2.3% accuracy on Tiny-ImageNet; in multi-modal learning, it demonstrates consistent gains of 4.1% accuracy on CMU-MOSEI and 2.8% accuracy on CMU-MOSI for sentiment analysis, while achieving 1.4% accuracy improvements on MISR. The code will be available on GitHub soon.