Bowen Cheng 程博文

Bowen is a third-year Ph.D. candidate in Electrical and Computer Engineering (ECE) at University of Illinois Urbana-Champaign (UIUC). His Ph.D. advisor is Prof. Thomas Huang and he is doing research in computer vision. Before commencing his graduate studies, he received his B.S. in ECE at UIUC in 2017.

Research Interests

I’m conducting research on various Computer Vision problems with Deep Learning as the tool. More specifically, my primary research goal is to solve multiple high-level image recognition problems (including image classification, object detection, semantic segmentation, instance/panoptic segmentation, human pose estimation, etc.) in a single unified pipeline. I’m also conducting research in neural architecture search (NAS).


  • [2019.11] Panoptic-DeepLab won the best paper, best result, and most innovative awards on the Mapillary Panoptic Segmentation track at ICCV 2019 Joint COCO and Mapillary Recognition Challenge Workshop.
  • [2019.10] Panoptic-DeepLab ranks first on Cityscapes panoptic and semantic segmentation benchmarks and third on instance segmentation benchmark with a single model (without finetuning on individual tasks).
  • [2019.07] One paper accepted in ICCV 2019.
  • [2019.07] One paper accepted in BMVC 2019.
  • [2018.10] One paper accepted to IEEE TIP.
  • [2018.10] We released code and models for our ECCV2018 paper with an extended tech report.
  • [2018.07] Two papers accepted in ECCV 2018.
  • [2017.10] MSRA team won 3rd place in COCO detection challenge 2017 by a simple and clean entry based on Deformable ConvNets (Slides)!
  • [2017.06] One paper accepted in ACII 2017.