Anomaly Localization
Anomaly localization aims to pinpoint atypical regions within data, such as defects in manufactured goods or lesions in medical images. Current research heavily utilizes deep learning, focusing on autoencoders, transformers, and generative models to learn normal patterns and identify deviations, often employing techniques like reconstruction error analysis, score matching, and knowledge distillation. This field is crucial for various applications, including industrial quality control, medical diagnosis, and security systems, offering the potential for improved efficiency, accuracy, and safety.
Papers
June 15, 2023
May 4, 2023
February 17, 2023
January 23, 2023
December 12, 2022
October 27, 2022
September 29, 2022
August 23, 2022
August 2, 2022
July 21, 2022
July 20, 2022
June 9, 2022
June 8, 2022
May 13, 2022
April 23, 2022
March 21, 2022
March 3, 2022
February 8, 2022
December 12, 2021