Yunfan Liu

Biography

I am an Associate Professor at the School of Advanced Interdisciplinary Sciences (SAIS), University of Chinese Academy of Sciences (UCAS). I am affiliated with the Learning and Machine Perception (LAMP) Lab, where I work with Prof. Qixiang Ye. Before joining SAIS, I was a postdoctoral researcher in the LAMP Lab.

I received my Ph.D. from UCAS in 2023, advised by Prof. Tieniu Tan. Before that, I received my M.S. from the University of Michigan - Ann Arbor in 2017 and my B.Eng. from Tsinghua University in 2015.

Research Interests

My research interests include multimodal representation and generative models, post-training techniques, and their applications to autonomous agents.

Recently, I have been particularly interested in how these models can support self-improving agentic systems.

[Opening Positions] I am actively looking for highly motivated and self-driven students. Prospective MS / Ph.D students with strong initiative and interests aligned with the topics above are welcome to contact me by email. RA positions are also available for highly motivated students (preferably undergraduate students as UCAS).

News

[2026-05] Our paper, "Balancing Understanding and Generation in Discrete Diffusion Models", has been accepted by ICML 2026 (Spotlight, 2.2%)!
[2026-03] I joined the School of Advanced Interdisciplinary Sciences at the University of Chinese Academy of Sciences!
[2026-02] Our paper, "RebRL: Reinforcing Discrete Visual Diffusion Models with Rebalanced Timestep Credits", has been accepted by CVPR 2026!
[2025-09] Our paper, "CC-Diff++: Spatially Controllable Text-to-Image Synthesis for Remote Sensing with Enhanced Contextual Coherence", has been accepted by TGRS!
[2025-08] Our paper, "Expandable Residual Approximation for Knowledge Distillation", has been accepted by TNNLS!
[2025-02] Two papers have been accepted by CVPR 2025, including one Highlight.

Selected Publications

^* Equal contribution. ^# Corresponding author.

Balancing Understanding and Generation in Discrete Diffusion Models

Yue Liu, Yuzhong Zhao, Zheyong Xie, Qixiang Ye, Jianbin Jiao, Yao Hu, Shaosheng Cao, Yunfan Liu^#

ICML 2026 (Spotlight, 2.2%)

[Paper] [Code]
RebRL: Reinforcing Discrete Visual Diffusion Models with Rebalanced Timestep Credits

Mu Zhang, Tianren Ma, Yunfan Liu^#, Kun Hu, Qixiang Ye

CVPR 2026

[Paper] [Code (Coming Soon)]
CC-Diff++: Spatially Controllable Text-to-Image Synthesis for Remote Sensing with Enhanced Contextual Coherence

Mu Zhang, Yunfan Liu^#, Yue Liu, Yuzhong Zhao, Qixiang Ye

TGRS 2025

[Paper] [Code]
Revealing Key Details to See Differences: A Novel Prototypical Perspective for Skeleton-based Action Recognition

Hongda Liu, Yunfan Liu^#, Min Ren, Hao Wang, Yunlong Wang, Zhenan Sun

CVPR 2025 (Highlight, 2.98%)

[Paper] [Code]
Building Vision Models upon Heat Conduction

Zhaozhi Wang^*, Yue Liu^*, Yunjie Tian, Yunfan Liu, Yaowei Wang, Qixiang Ye

CVPR 2025

[Paper] [Code]
VMamba: Visual State Space Model

Yue Liu, Yunjie Tian, Yuzhong Zhao, Hongtian Yu, Lingxi Xie, Yaowei Wang, Qixiang Ye, Yunfan Liu^#

NeurIPS 2024 (Spotlight, 2.08%)

[Paper] [Code]
Spatial Transform Decoupling for Oriented Object Detection

Hongtian Yu^*, Yunjie Tian^*, Qixiang Ye, Yunfan Liu^#

AAAI 2024

[Paper] [Code]
GAN-based Facial Attribute Manipulation

Yunfan Liu, Qi Li, Qiyao Deng, Zhenan Sun, Ming-Hsuan Yang

TPAMI 2023

[Paper]
Towards Spatially Disentangled Manipulation of Face Images With Pre-Trained StyleGANs

Yunfan Liu, Qi Li, Qiyao Deng, Zhenan Sun

TCSVT 2022

[Paper]
A³GAN: An Attribute-Aware Attentive Generative Adversarial Network for Face Aging

Yunfan Liu, Qi Li, Zhenan Sun, Tieniu Tan

TIFS 2021

[Paper]
Age Progression and Regression with Spatial Attention Modules

Qi Li^*, Yunfan Liu^*, Zhenan Sun

AAAI 2020 (Oral)

[Paper] [Code]
Attribute-Aware Face Aging With Wavelet-Based Generative Adversarial Networks

Yunfan Liu^*, Qi Li^*, Zhenan Sun

CVPR 2019

[Paper]