Zhihao Duan

PhD candidate at Purdue University

prof_pic.jpg

Hi! I’m a PhD student at Purdue University ECE, the Video and Image Processing Lab. My research lies in the intersection of data compression, computer vision, and machine learning. Our goal is to use deep learning to improve visual data compression (compression with AI) and design compressors tailored to downstream vision tasks (compression for AI). I also work with colleagues on neural network model compression (compression of AI).

Education: I received my Bachelor’s degree in Electrical Engineering from Shanghai Jiao Tong University, China, in 2018. I received my Master’s degree in ECE from Boston University, USA, in 2020. I’m currently pursuing my PhD at Purdue University, USA, and expect to graduate in late 2024.

news

Feb 27, 2024 Our paper “Towards Backward-Compatible Continual Learning of Image Compression” is accepted to CVPR 2024!
Oct 04, 2023 Our paper “QARV: Quantization-Aware ResNet VAE for Lossy Image Compression” is accepted to IEEE TPAMI!
Jan 04, 2023 Our paper “Lossy Image Compression with Quantized Hierarchical VAEs” is awarded the Best Algorithms Paper Award at WACV 2023!

selected publications

  1. continual-compression.gif
    Towards Backward-Compatible Continual Learning of Image Compression
    Zhihao Duan, Ming Lu , Justin Yang , Jiangpeng He , Zhan Ma , and Fengqing Zhu
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Jun 2024
  2. qarv.gif
    QARV: Quantization-Aware ResNet VAE for Lossy Image Compression
    Zhihao Duan, Ming Lu , Jack Ma , Yuning Huang , Zhan Ma , and Fengqing Zhu
    IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), Jun 2024
  3. qres.gif
    Lossy Image Compression with Quantized Hierarchical VAEs
    Zhihao Duan, Ming Lu , Zhan Ma , and Fengqing Zhu
    IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), Jan 2023
    Best Paper Award - Algorithms
  4. feature-compression.gif
    Efficient Feature Compression for Edge-Cloud Systems
    Zhihao Duan, and Fengqing Zhu
    Picture Coding Symposium, Dec 2022
    Best Paper Award Finalists (top 5%)
  5. rapid.gif
    RAPiD: Rotation-Aware People Detection in Overhead Fisheye Images
    Zhihao Duan, M. Ozan Tezcan , Hayato Nakamura , Prakash Ishwar , and Janusz Konrad
    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, Jun 2020