Found in Translation: semantic approaches for enhancing AI interpretability in face verification
This work addresses the need for human-centric interpretability in critical applications like face verification, though it is incremental as it extends previous methods with semantic features.
The study tackled the problem of low interpretability in face verification AI by integrating semantic concepts from human cognition into explainable AI frameworks, resulting in user preference for semantic explanations over traditional heatmaps and improved nuanced understanding of model decisions.
The increasing complexity of machine learning models in computer vision, particularly in face verification, requires the development of explainable artificial intelligence (XAI) to enhance interpretability and transparency. This study extends previous work by integrating semantic concepts derived from human cognitive processes into XAI frameworks to bridge the comprehension gap between model outputs and human understanding. We propose a novel approach combining global and local explanations, using semantic features defined by user-selected facial landmarks to generate similarity maps and textual explanations via large language models (LLMs). The methodology was validated through quantitative experiments and user feedback, demonstrating improved interpretability. Results indicate that our semantic-based approach, particularly the most detailed set, offers a more nuanced understanding of model decisions than traditional methods. User studies highlight a preference for our semantic explanations over traditional pixelbased heatmaps, emphasizing the benefits of human-centric interpretability in AI. This work contributes to the ongoing efforts to create XAI frameworks that align AI models behaviour with human cognitive processes, fostering trust and acceptance in critical applications.