Explainable Machine Learning: A Multi-Modal Approach to Human-AI Collaboration

Trust is paramount in high-stakes AI applications. This research introduces a multi-modal explainability framework that provides visual, textual, and saliency-based explanations for model predictions. Through extensive user studies, we show that our approach significantly enhances user trust and performance in collaborative decision-making tasks.