Hi there Welcome to my Homepage!

Hi! I am Qifan Liang (Simon Leong), a first-year master student (MComp AI) at the Sound and Music Computing Lab, National University of Singapore, supervised by Prof. Ye Wang. Previously, I conducted computer vision research under Prof. Jens Rittscher at Oxford University and Prof. Zhen Han, Prof. Zhongyuan Wang at National Multimedia Software Engineering Technology Research Center, Wuhan University.

Feel free to reach out if you are interested in collaboration or potential opportunities!

“Computer vision will open our eyes to a new world of possibilities, where machines not only see but understand the world as we do.” — Inspired by Fei-Fei Li

• Controllable generative modeling in Audio-Visual Multimodal Learning.
• Low- and high-level computer vision for effective visual representation and understanding.
• Foundations and applications of human-centric perception and 3D scene understanding.

National University of Singapore - SMC Lab
2025.11 – Present
Master of Computing (AI) at Sound and Music Computing Lab, supervised by Prof. Ye Wang

Oxford University – Big Data Institute
2024.06 – 2024.10
Remote Summer Research Intern at Oxford Big Data Institute, supervised by Prof. Jens Rittscher

Wuhan University – Multimedia Lab
2023.05 – 2025.06
Research Assistant at National Multimedia Software Engineering Technology Research Center, supervised by Prof. Zhen Han

Core Publications Full Publications List

(* corresponding author  ·  † equal contribution)

PersonaGest: Personalized Co-Speech Gesture Generation with Semantic-Guided Hierarchical Motion Representation
Junchuan Zhao^†, Qifan Liang^†, Ye Wang*
A two-stage framework for personalized co-speech gesture generation that disentangles content from style via a semantic-guided RVQ-VAE, then uses masked generative and style residual transformers to produce semantically coherent, style-consistent gestures with fine-grained part-wise control.
arXiv 2026   [arXiv] [website]

TED-TTS: Training-Free Intra-Utterance Emotion and Duration Control for Text-to-Speech Synthesis
Qifan Liang^†, Y Liu^†, R Wei^†, N Lu, J Zhao*, Ye Wang
A training-free framework for fine-grained intra-utterance emotion and duration control in TTS, using segment-aware emotion and duration conditioning with causal masking and monotonic stream alignment filtering.
ACL 2026   🎙 Oral Nominee (SAC) [arXiv] [website] [code]

Rethinking Surgical Smoke: A Smoke-Type-Aware Laparoscopic Video Desmoking Method and Dataset
Qifan Liang, Junlin Li, Zhen Han*, Xihao Wang, Zhongyuan Wang, Bin Mei
A smoke-type-aware network (STANet) for laparoscopic video desmoking, and introduce STSVD, a large-scale synthetic dataset with smoke type annotations across 28 surgical scenarios.
AAAI 2026   [arXiv] [website] [code]

Few-Shot Face Sketch-to-Photo Synthesis via Global-Local Asymmetric Image-to-Image Translation
Yongkang Li, Qifan Liang, Wenjun Mai, Zhen Han*, Zhongyuan Wang
A global-local asymmetric translation framework for few-shot face sketch-to-photo synthesis, combining style transfer and structural alignment to produce realistic photo-realistic outputs.
TOMM 2024   [paper]

Multi-Source Perception and Intelligent Prediction Emergency Monitoring Platform for Forest Fires
Qifan Liang (Team Lead)
An intelligent forest fire monitoring platform integrating multi-source perception, real-time prediction, and emergency response.
Project   [code]
🏆 National Grand Prize (Top 0.5%) · 2023 National University Student Surveying & Mapping Competition
🏆 National First Prize (Top 1.7%) · 2023 "Dingxin Cup" National Youth Innovation Competition
🏆 National Second Prize (Top 0.7%) · 2023 Chinese Collegiate Computing Competition

• 2025.05 Outstanding Graduation Thesis · Wuhan University
• 2025.05 Lei Jun Computer Innovation and Development Fund & Lei Jun Computer Research Grant (Rate: 7.2%, major-wide) · Xiaomi Corporation
• 2025.03 Outstanding Graduate · School of Computer Science, Wuhan University
• 2024.11 Outstanding Undergraduate Scholarship (10/685, major-wide) · DiDi Inc.
• 2024, 2023.10 Merit Student (Top 5%, school-wide) · Wuhan University
• 2024.05 National University Students’ Innovation & Entrepreneurship Fund (Top 3%, school-wide) · Wuhan University
• 2024.03 Lei Jun Computer Innovation and Development Fund (Rate: 7.2%, major-wide) · Xiaomi Corporation
• 2023.11 Outstanding Undergraduate Scholarship (1/685, major-wide) · Samsung Group
• 2023.10 First-Class Scholarship (Top 5%, school-wide) · Wuhan University
• 2023.06 National Grand Prize (Top 0.5%, nation-wide) · 2023 National University Student Surveying & Mapping Competition
• 2022.10 Outstanding Student (Top 30%, school-wide) · Wuhan University

• Reviewer, AAAI Conference on Artificial Intelligence (AAAI 2026)
• Poster Presenter, AAAI Conference on Artificial Intelligence (AAAI 2026)
• Presenter, Annual Meeting of the Association for Computational Linguistics (ACL 2026)
• Conference Participant, China Computer Federation (CCF) Computer Networking Conference, Wenzhou, China, 2023
• Conference Participant, Intelligent Transportation Networking and Telematics Yongjia Forum, Yongjia, China, 2023