Kang-Yang Huang

About Me

Kang-Yang (Albert Huang) obtained his MSc degree in Computer Science at National Taiwan University, Taiwan, where he is part of the AIMM group within the Communication and Multimedia Laboratory (CMLab), under the guidance of Prof. Wen-Huang Cheng and Prof. Hong-Han Shuai. Albert previously earned his B.S. degree from the Undergraduate Honors Program in Electrical Engineering and Computer Science at National Yang Ming Chiao Tung University, Taiwan. Additionally, he was an exchange student at RWTH Aachen University, Germany, from April to September in 2021. Albert’s research interests are centered around the convergence of computer vision and deep learning. In the early stages of his academic journey, Albert specialized in lightweight deep learning, particularly in the areas of knowledge distillation and federated learning. Currently, Albert is broadening his research scope to include text-to-image generation and visual language models.

Research Interests

• Computer Vision: image generation, image editing, video anomaly detection, interpretable DeepFake detection
• Deep Learning: federated learning, knowledge distillation, adversarial attack

News

• [Dec. 2025] We releaase Tri-DF, the first multimodal benchmark designed for interpretable DeepFake detection on arXiv.
• [Jul. 2025] Our research paper about face anti-spoofing utilizing vision-language model is accepted by ACM MM 2025.
• [Jun. 2025] Member of the Phi Tau Phi Scholastic Honor Society, 2025.
• [Jun. 2025] Received Hon Hai Technology Award, 2025.
• [Jan. 2025] Our survey paper about lightweight deep learning received the Best Paper Award for the year of 2024.
• [Nov. 2024] Received Lam Research Paper Award and Outstanding Student Scholarship, 2024.
• [Jul. 2024] Our research paper about image generation with precise human-object interactions through intelligent reasoning and correction is accepted by ACM MM 2024.
• [Jun. 2024] Our research paper about weakly-supervised anomaly detection is accepted by IJCAI Workshop 2024.
• [Mar. 2024] Our research paper about HOI detection is accepted for Oral by ICME 2024.
• [Jun. 2022] Our survey paper about lightweight deep learning is accepted by IEEE Consumer Electronics Magazine.
• [Jul. 2021] Our research paper about federated learning is accepted for Poster by ICME 2021.

Publications

Powered by Jekyll and Minimal Light theme.