Transparent Visual Reasoning via Object-Centric Agent Collaboration

2025-09-28

AI Summary

The core challenge in interpretable visual reasoning lies in generating transparent decision processes grounded in human-understandable concepts. To address this, we propose OCEAN—a novel framework featuring a game-theoretic, multi-agent negotiation mechanism integrated with end-to-end learned object-centric representations, enabling intrinsically interpretable visual reasoning. Each agent models individual scene objects and jointly negotiates to produce coherent, human-identifiable reasoning evidence—without post-hoc processing. On multi-object benchmarks, OCEAN achieves accuracy competitive with state-of-the-art black-box models. User studies demonstrate that its explanations are significantly more intuitive and trustworthy than post-hoc methods (e.g., Grad-CAM, LIME), yielding substantial gains in both faithfulness and comprehensibility. This work establishes a new paradigm for explainable AI that harmonizes theoretical rigor—rooted in game-theoretic principles—with cognitive alignment to human reasoning.

Technology Category

• Computer Vision: Interpretability, Explainability, and Transparency
• Humans and AI: Explainable AI (XAI) for Human Understanding
• Knowledge Representation and Reasoning: Common-Sense Reasoning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

A central challenge in explainable AI, particularly in the visual domain, is producing explanations grounded in human-understandable concepts. To tackle this, we introduce OCEAN (Object-Centric Explananda via Agent Negotiation), a novel, inherently interpretable framework built on object-centric representations and a transparent multi-agent reasoning process. The game-theoretic reasoning process drives agents to agree on coherent and discriminative evidence, resulting in a faithful and interpretable decision-making process. We train OCEAN end-to-end and benchmark it against standard visual classifiers and popular posthoc explanation tools like GradCAM and LIME across two diagnostic multi-object datasets. Our results demonstrate competitive performance with respect to state-of-the-art black-box models with a faithful reasoning process, which was reflected by our user study, where participants consistently rated OCEAN's explanations as more intuitive and trustworthy.

Problem

• Developing explainable AI for visual reasoning using human-understandable concepts
• Creating interpretable object-centric frameworks with transparent multi-agent collaboration
• Producing faithful visual explanations that outperform posthoc methods like GradCAM

Innovation

• Object-centric representations for visual reasoning
• Multi-agent negotiation for transparent decision-making
• Game-theoretic approach for coherent evidence selection