About Me

I received my Bachelor's degree from Guangdong University of Technology, where my early explorations in web-supervised fine-grained image recognition led to a granted national patent. Currently, I am a Master's student at South China University of Technology (SCUT), privileged to be advised by Prof. Tianshui Chen and Prof. Chunmei Qing.

Driven by a deep curiosity about AI, my earlier research centered on Generative AI and Controllable Image Synthesis. My work on Speech-Preserving Facial Expression Manipulation systematically addressed data scarcity via representation learning, resulting in top-tier publications including a CVPR 2024 Highlight, T-PAMI, and IJCV. Concurrently, my exploration of Diffusion-based Editing resolved the trade-off between editability and visual consistency, culminating in a sole first-author paper in TIP 2025 and another pending patent.

Currently, driven by a strong desire to ground AI in the physical world, my core research focus has transitioned to Embodied AI. Embracing an Industry-Academia Fusion paradigm, I am actively engineering the end-to-end robotic pipeline, from multi-arm hardware setup and teleoperation to VLA model training. Building upon this hands-on experience, I am now researching architectural solutions for critical real-world deployment failures (e.g., representation degradation), aspiring to distill practical engineering bottlenecks into my next top-tier academic contributions.

News

• 2026.04
  STCCL was accepted by T-PAMI.
• 2025.09
  Neural Scene Designer was published in IEEE TIP.
• 2025.08
  CDRL was published in IJCV.
• 2024.09
  Started M.Eng. study at SCUT.
• 2024.06
  ASCCL was selected as a CVPR 2024 Highlight (Top 2.8%).
• 2024.01
  Joined X-Era AI as an algorithm research intern.

Selected Publications

Controllable Image Editing

Neural Scene Designer: Self-Styled Semantic Image Manipulation

Jianman Lin, Tianshui Chen, Chunmei Qing, Zhijing Yang, Shuangping Huang, Yuheng Ren, Liang Lin.

IEEE Transactions on Image Processing (TIP), 2025 · CCF-A PDF Page

Geometry-Editable and Appearance-Preserving Object Composition

Jianman Lin, Haojie Li, Chunmei Qing, Zhijing Yang, Liang Lin, Tianshui Chen.

Under review · IJCV (CCF-A) PDF Page

Speech-Preserving Facial Expression Manipulation

Learning Spatial-Temporal Coherent Correlations for Speech-Preserving Facial Expression Manipulation

Jianman Lin, Tianshui Chen, Zhijing Yang, Chunmei Qing, Liang Lin.

IEEE Transactions on Pattern Analysis and Machine Intelligence (T-PAMI), 2026 · CCF-A PDF Project

Learning Adaptive Spatial Coherent Correlations for Speech-Preserving Facial Expression Manipulation

Tianshui Chen, Jianman Lin, Zhijing Yang, Chunmei Qing, Liang Lin.

CVPR 2024 · Highlight (Top 2.8%) · CCF-A PDF Page

Contrastive Decoupled Representation Learning and Regularization for Speech-Preserving Facial Expression Manipulation

Tianshui Chen, Jianman Lin*, Zhijing Yang, Chunmei Qing, Yukai Shi, Liang Lin.

International Journal of Computer Vision (IJCV), 2025 · CCF-A PDF Page

AI4Science & Interdisciplinary

Multiscale Exploration of Informative Latent Features for Accurate Deep Eutectic Solvents Viscosity Prediction

Jianman Lin, et al.

AIChE Journal, 2025 · JCR Q1 DOI Page

* Equal contribution.

Honors & Awards

• CVPR 2024 Highlight — Top 2.8% of accepted papers.
• First Prize (National), China Undergraduate Mathematical Contest in Modeling (CUMCM), sponsored by Higher Education Press.
• Third Prize (National), National Mechanical Engineering Innovation and Creativity Competition.

Patents

• Granted
  Invention patent on web-supervised fine-grained image recognition. National invention patent (China).
• Granted
  Method for diffusion-based controllable image editing. National invention patent (China).

Academic Service

• Conference Reviewer: CVPR, ECCV, NeurIPS.

Experience

Education