The content discusses the importance of robust and explainable models in computer vision for trustworthy AI applications. It covers the challenges faced in deploying machine learning and deep learning models in practice, focusing on robustness, explainability, and reliability. The thesis explores the development of interpretable classifiers using radial basis function networks (RBFs) for CNNs, the detection of adversarial attacks using feature response maps, and the reduction of motion artifacts in medical imaging. It also delves into automated machine learning and deep learning, showcasing successful applications in affective computing, medical imaging, and fairness in face recognition systems.
Sang ngôn ngữ khác
từ nội dung nguồn
arxiv.org
Thông tin chi tiết chính được chắt lọc từ
by Mohammadreza... lúc arxiv.org 03-28-2024
https://arxiv.org/pdf/2403.18674.pdfYêu cầu sâu hơn