I'm Ziyun Yang,
a Computer Vision Researcher.
Welcome
ML Researcher at Apple & PhD from Duke University.
Biography
Biography
I'm Ziyun Yang, a researcher in ML.
I am a ML research engineer at Apple, working on low-level vision and compression. Before that, I obtained my Ph.D. from Biomedical Engineering at Duke, supervised by Dr. Sina Farsiu .
I develop innovating cutting-edge computer vision techniques on several cross-discipline problems, including salient&camouflaged object detection, semantic segmentation on natural & biomedical images, generative model-based image denoising & synthesis, and multi-model classification.
My goal is to increase the performance of artificial intelligence with minimum computational costs. So far, my works have been published in TPAMI, CVPR, Pattern Recognition, IEEE TMI, and Biomedical Optic Express, etc.
- Name:Ziyun Yang
- Email:ziyun.yang@duke.edu
- Phone:919-884-0064
- Social: Google Scholar , LinkedIn
Keywords
Research Interests
Machine & Deep Learning
Develop reliable and high-performance machine learning & and deep learning methods. Develop connectivity-based deep learning framework for computer vision applications.
Semantic Segmentation
Develop SOTA Segmentation Model for various applications, including natural image segmentation, biomedical image analysis etc..
Salient & Camouflaged Object Detection
Detecting the saliency and the concealed objects in a given scene .
Medical Imaging & Analysis
3D reconstruction and biomedical image analysis (i.e., segmentation, registration, diagnosis etc.).
Image Synthesis
Develop Generative Learning-based methodology in image systhesis for downstream tasks.
Learned Image Compression
Work on learned image compression and other low-level tasks.
Summary
Experience
Education
2019 - 2024
Duke University
Ph.D. in Biomedical Engineering
Expected graduation: Sep 2024. Advisor: Dr. Sina Farsiu
2015 - 2019
Beijing Institute of Technology
Bachelor of Science in Automation Engineering
• Thesis title: Multimodal Neural Network for Data Fusion in Image Classification.
Work Experience
November 2024 - Current
Apple
Machine Learning Research Engineer
Working on learned video/image compression, super resolution, and other low-level vision problems.
May 2023 - August 2023
Meta
Research Scientist Intern – Computer Vision & Eye Tracking
Developed segmentation and registration algorithms for improving 3D reconstruction for eye tracking research in a multi-disciplinary team. Developed software to facilitate calibration process for eye tracking device.
July 2018 - Oct 2018
Johns Hopkins University
Research Intern
Applied reverse adversarial training in polyp detection, supervised by Dr. Faisal Mahmood and Dr. Nicholas Durr .
Portfolio
Representative Work
Publications
Publications
- Spatial Coherence Loss for Salient and Camouflaged Object Detection and Beyond
Ziyun Yang, Kevin Choy, and Sina Farsiu
Pattern Recognition, vol 173, p. 112798, 2026 - Segment concealed objects with incomplete supervision
Chunming He, Kai Li, Yachao Zhang, Ziyun Yang, Youwei Pang, Longxiang Tang, Chengyu Fang, Yulun Zhang, Linghe Kong, Xiu Li, Sina Farsiu
IEEE TPAMI, vol. 47, no. 9, pp. 7832-7851, 2025 - Self-knowledge distillation-empowered directional connectivity transformer for microbial keratitis biomarkers segmentation on slit-lamp photography
Ziyun Yang, Maria A Woodward, Leslie M Niziol, Mercy Pawar, N Venkatesh Prajna, Anusha Krishnamoorthy, Yiqing Wang, Ming-Chen Lu, Suvitha Selvaraj, Sina Farsiu
Medical Image AnalysisVolume 102, 103533, 2025 - Directional Connectivity-based Segmentation of Medical Images
Ziyun Yang and Sina Farsiu
CVPR 2023, pp. 2774-2784 - BiconNet: An Edge-preserved Connectivity-based Approach for Salient Object Detection
Ziyun Yang, Somayyeh Soltanian-Zadeh, and Sina Farsiu
Pattern Recognition, vol. 121, p. 108231, 2022. - RetiFluidNet: A Self-Adaptive and Multi-Attention Deep Convolutional Network for Retinal OCT Fluid Segmentation
R. Rasti, A. Biglari, M. Rezapourian, Z. Yang and S. Farsiu
IEEE Transaction on Medical Imaging, vol. 42, no. 5, pp. 1413-1423, 2023. - Connectivity-based deep learning approach for segmentation of the epithelium in in vivo human esophageal OCT images
Ziyun Yang, Somayyeh Soltanian-Zadeh, Kengyeh K. Chu, Haoran Zhang, Lama Moussa, Ariel E. Watts, Nicholas J. Shaheen, Adam Wax, and Sina Farsiu
Biomedical Optics Express, vol. 12, no. 10, pp. 6326-6340, 2021. - Multimodal coherent imaging of retinal biomarkers of Alzheimer’s disease in a mouse model
Ge Song, Zachary A Steelman, Stella Finkelstein, Ziyun Yang, Ludovic Martin, Kengyeh K Chu, Sina Farsiu, Vadim Y Arshavsky, Adam Wax
Scientific Reports 10, 7912, 2020 - Fusing Attributes Predicted via Conditional GANs for Improved Skin Lesion Classification
F. Mahmood, J. Johnson, Z. Yang, N. J. Durr
Proc. SPIE 10950, 109501T, 2019. - Polyp Segmentation and Classification using Predicted Depth from Monocular Endoscopy
F. Mahmood, Z. Yang, R. Chen, D. Borders, W. Xu, N. J. Durr
Proc. SPIE 10950, 1095011, 2019. - Multimodal Densenet
F. Mahmood, Z. Yang, T. Ashley, N. J. Durr
arXiv:1811.07407, 2018.