Dr. Xinyu Yang
Postdoc
Xinyu Yang is a Senior Research Associate at Lancaster University, specializing in the development of innovative AI model training and learning technologies. With a background in Machine Learning, Deep Learning, and Computer Vision, Xinyu's research focuses on reducing the need for manual labeling in various computer vision tasks. He holds a PhD in Computer Vision from the University of Bristol, where his research concentrated on the application of AI to animal-related fields. His research interests include biometrics, image synthesis, and advanced learning paradigms such as self/weakly-supervised and zero-shot learning, with relevant work published in several conferences and journals.
Email: x.yang28 (at) lancaster.ac.uk
Office: InfoLab21 A25, School of Computing and Communications, Lancaster University
Research
Weakly Supervised Co-training with Swapping Assignments for Semantic Segmentation
Xinyu Yang, Hossein Rahmani, Sue Black, Bryan M. Williams ECCV, 2024 [Webpage] [arXiv] [Code] [pdf] |
Dynamic Curriculum Learning for Great Ape Detection in the Wild
Xinyu Yang, Tilo Burghardt, Majid Mirmehdi IJCV, 2023 [Webpage] [arXiv] [Code] [pdf] |
Back to the Future: Cycle Encoding Prediction for Self-supervised Video Representation Learning
Xinyu Yang, Majid Mirmehdi, Tilo Burghardt BMVA British Machine Vision Conference, 2021 [Webpage] [arXiv] [Code] [pdf] |
Great Ape Detection in Challenging Jungle Camera Trap Footage via Attention-Based Spatial and Temporal Feature Blending
Xinyu Yang, Majid Mirmehdi, Tilo Burghart Computer Vision for Wildlife Conservation on ICCV, 2019 [Webpage] [arXiv] [Code] [pdf] |
Misc.
Teaching
- COMSM0045 Applied Deep Learning
- COMS30030 Image Processing and Computer Vision
- COMS20011 Data-Driven Computer Science
Reviewer
- CVPR2022
- BMVC2021
- IET Computer Vision Jouneral 2019