Review paper tracks evolution of visual explanation methods in AI

A new review paper published on arXiv details the evolution of Class Activation Mapping (CAM) techniques in explainable computer vision. The paper categorizes 57 method-centered studies from 2016 onwards, highlighting a shift from explaining simple CNN classifiers to more complex methods involving transformers, foundation models like CLIP and DINO, and attention mechanisms. It notes that while explanation methods are becoming more sophisticated, evaluation protocols for faithfulness, robustness, and human trust remain fragmented. AI

IMPACT Provides a structured overview of visual explanation techniques, aiding researchers in understanding the landscape and identifying evaluation gaps.

RANK_REASON The cluster contains a single academic paper published on arXiv, detailing a review of existing research methods. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• AmirHossein Eshghi
• arXiv
• Class activation mapping
• CNN
• DINO
• Explainable Computer Vision
• foundation model
• SAM
• Transformer++

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →