TATL: Task Agnostic Transfer Learning for skin attributes detection

• Main Authors: NGUYEN, Duy M.H., NGUYEN, Thu T., VU, Huong, PHAM, Hong Quang, NGUYEN, Manh-Duy, NGUYEN, Binh T., SONNTAG, Daniel
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2022
• Subjects: Encoder-decoder architecture; skin attribute detection; transfer learning; Artificial Intelligence and Robotics; Health Information Technology
• Online Access: https://ink.library.smu.edu.sg/sis_research/7825; https://ink.library.smu.edu.sg/context/sis_research/article/8828/viewcontent/TATL_av.pdf
• Institution: Singapore Management University

Existing skin attributes detection methods usually initialize with a pre-trained Imagenet network and then fine-tune on a medical target task. However, we argue that such approaches are suboptimal because medical datasets are largely different from ImageNet and often contain limited training samples...