Weixia Zhang (张维夏)
Associate Research Scientist (副研究员)
Institute of AI, School of Computer Science, Shanghai Jiao Tong University
Welcome to MIGA Lab. We explore how visual content can be measured, improved, captured, and understood. Our research spans multimedia quality assessment, perceptual optimization, computational photography, and task-driven computer vision. We develop perceptual and task-aware models to understand what makes an image “great”, use them to guide visual enhancement, improve image quality at the moment of capture, and extend visual intelligence to real-world applications. Ultimately, we aim to make images not only look better, but also work better—for both humans and machines.
Multimedia Quality Assessment
Focusing on accurate and fair evaluation of multimedia quality, while also serving as loss functions, reward models, or real-time feedback signal in shaping various vision systems.
Perceptual Optimization
Working on existing visual content, seeking to enhance, restore, or edit images/videos to improve their perceived quality.
Computational Photography
Targeting the visual capture stage, aiming to produce high-quality images/videos directly at the moment of acquisition.
Practical Vision Analysis
Beyond quality assessment for human vision system, we aim to develop vision systems for real-world applications such as industrial safety surveillance.
See the full publications list for more details.