Chenhongyi Yang 杨陈弘毅
pdb

I am a PhD student at the School of Engineering, University of Edinburgh. My principal supervisor is Dr. Elliot J. Crowley. I am also a member of the BayesWatch group.

My research is mainly about computer vision, and my research interests include:

  • Developing image recognition algorithms under insufficient human annotations
  • Learning good visual representaiton in an unsupervised way
  • Designing high-performance neural network architectures

Previously, I did my MSc in computer science at Boston University and my BEng in computer science at the University of Science and Technology of China.

Conference Papers

Prediction-Guided Distillation for Dense Object Detection

ECCV 2022

Chenhongyi Yang, Mateusz Ochal, Amos Storkey, Elliot J. Crowley.

PGD is a high-performing knowledge distillation framework designed for single-stage object detectors. It distills every object in a few key predictive regions and uses an adaptive weighting scheme to compute the foreground feature imitation loss in those regions.

QueryDet: Cascaded Sparse Query for Accelerating High-Resolution Small Object Detection

CVPR 2022 (Oral)

Chenhongyi Yang, Zehao Huang, Naiyan Wang.

QueryDet achieves fast and accurate small object detection. Its core is the cascaded sparse query mechanism: rough locations of small objects are first found on low-resolution features, then those objects are accurately detected on high-resolution features using the efficient sparse convolution.

Disentangle Your Dense Object Detector

ACM Multimedia 2021 (Oral)

Zehui Chen*, Chenhongyi Yang*, Qiaofei Li, Feng Zhao, Zheng-Jun Zha, Feng Wu.

We investigated the conjunction problem in the modern dense object detectors, based on which we proposed the Disentangled Dense Object Detector (DDOD) where three effective disentanglement mechanisms were designed for boosting dense object detectors' performance.

Consistency Regularization with High-dimensional Non-adversarial Source-guided Perturbation for Unsupervised Domain Adaptation in Segmentation

AAAI 2021

Kaihong Wang, Chenhongyi Yang, Margrit Betke.

BiSIDA is a bidirectional style-induced domain adaptation method that employs consistency regularization to exploit information from the unlabeled target domain dataset. BiSIDA is easy to train because the domain adaptation is achieved with a simple nerual style transfer model.

Learning to Separate: Detecting Heavily-Occluded Objects in Urban Scenes

ECCV 2020

Chenhongyi Yang, Vitaly Ablavsky, Kaihong Wang, Qi Feng, Margrit Betke.

SG-NMS is a new non-maximum-suppression algorithm designed for detecting heavily-occluded objects. It is based on the semantic-geometry embedding mechanism where the embeddings of boxes belonging to the same object are pulled together and embeddings of boxes belong to different objects are pushed away. Then NMS is conducted based on the embedding distances.

Preprint Papers

Plug and Play Active Learning for Object Detection

Arxiv Preprint, Nov. 2022

Chenhongyi Yang, Lichao Huang, Elliot J. Crowley.

PPAL is a plug-and-play active learning framework for object detection. It is based on two innotations: a novel object-level uncertainty re-weighting mechanism and a new similarity computing method for designed multi-instance images.

DETRDistill: A Universal Knowledge Distillation Framework for DETR-families

Arxiv Preprint, Nov. 2022

Jiahao Chang*, Shuo Wang*, Guangkai Xu*, Zehui Chen, Chenhongyi Yang, Feng Zhao

DETRDistill is a knolwedge distillation framework designed for the DETR family, the transformer-based object detection architectures. The distillation is conducted in three parts: instance query distillation, visual feature distillation and bipartite matching distillation.

Contrastive Object-level Pre-training with Spatial Noise Curriculum Learning

Arxiv Preprint, Nov. 2021

Chenhongyi Yang, Lichao Huang, Elliot J. Crowley.

CCOP is an object-level self-supervised learning framework. It is based on contrasting the regional features of rough object boxes, which are found using an unsupervised way. We also develop a curriculum learning mechanism to alleviate the gradient vanishing problem.

Latest update:

Thanks to Jack Turner for the website template.