Fang Zhao

I am an Associate Professor at the School of Intelligence Science and Technology, Nanjing University. My research is currently on visual understanding in an open world, transfer learning and anomaly detection. Previously, I was a senior researcher in Tencent AI Lab from 2021 to 2023, and a research scientist in Inception Institute of Artificial Intelligence (IIAI) from 2018 to 2021, where I worked with Prof. Shengcai Liao. I was a research fellow in National University of Singapore (NUS) from 2015 to 2017, co-supervised by Prof. Jiashi Feng and Prof. Shuicheng Yan. I received my Ph.D. degree from Institute of Automation, Chinese Academy of Sciences (CASIA) under the supervision of Prof. Liang Wang.

Email  /  Google Scholar  /  Github

profile photo
Research
Attentive WaveBlock: Complementarity-enhanced Mutual Networks for Unsupervised Domain Adaptation in Person Re-identification and Beyond
Wenhao Wang, Fang Zhao, Shengcai Liao, Ling Shao
TIP, 2022
arXiv / Code / bibtex

This paper proposes a novel light-weight module, the Attentive WaveBlock (AWB), which can be integrated into the dual networks of mutual learning to enhance the complementarity.

Unsupervised Domain Adaptation with Noise Resistible Mutual-Training for Person Re-identification
Fang Zhao, Shengcai Liao, Guo-Sen Xie, Jian Zhao, Kaihao Zhang, Ling Shao
ECCV, 2020
bibtex

A Noise Resistible Mutual-Training (NRMT) method, which maintains two networks during training to perform collaborative clustering and mutual instance selection.

Dynamic Conditional Networks for Few-Shot Learning
Fang Zhao*, Jian Zhao*, Shuicheng Yan, Jiashi Feng (* - equal contribution)
ECCV, 2018
Code / bibtex

A novel Dynamic Conditional Convolutional Network (DCCN) is proposed to handle conditional few-shot learning, i.e, only a few training samples are available for each condition.

Learning Anchor Transformations for 3D Garment Animation
Fang Zhao, Zekun Li, Shaoli Huang, Junwu Weng, Tianfei Zhou, Guo-Sen Xie, Jue Wang, Ying Shan
CVPR, 2023
arXiv / Project / bibtex

This paper proposes an anchor-based deformation model, namely AnchorDEF, to predict 3D garment animation from a body motion sequence.

Learning Anchored Unsigned Distance Functions with Gradient Direction Alignment for Single-view Garment Reconstruction
Fang Zhao, Wenhao Wang, Shengcai Liao, Ling Shao
ICCV, 2021 (Oral)
arXiv / Code / bibtex

We propose a novel learnable Anchored Unsigned Distance Function (AnchorUDF) representation for 3D garment reconstruction from a single image.

Human Parsing Based Texture Transfer from Single Image to 3D Human via Cross-View Consistency
Fang Zhao, Shengcai Liao, Kaihao Zhang, Ling Shao
NeurIPS, 2020
Code / bibtex

A human parsing based texture transfer model via cross-view consistency learning which generates the texture of 3D human body from a single image.

Robust LSTM-Autoencoders for Face De-Occlusion in the Wild
Fang Zhao, Jiashi Feng, Jian Zhao, Wenhan Yang, Shuicheng Yan
TIP, 2018
arXiv / Code / bibtex

We propose a robust LSTM-Autoencoders (RLA) model consisting of two LSTM components, which aims at occlusion-robust face encoding and recurrent occlusion removal respectively.

Deep Semantic Ranking Based Hashing for Multi-Label Image Retrieval
Fang Zhao, Yongzhen Huang, Liang Wang, Tieniu Tan
CVPR, 2015
arXiv / Code / bibtex

We propose a deep semantic ranking based method for learning hash functions that preserve multilevel semantic similarity between multi-label images.