I'm an undergraduate researcher at Jiangnan University (Project 211 Β· Double First-Class), School of AI & Computer Science, B.Eng. in Digital Media Technology β Class of 2027 (Junior) Β· GPA 88 / 100.
My research spans Computer Vision (medical image segmentation & vision-language fusion), AI4Science (protein & small-molecule design), and Large Language Models (reasoning & neuro-symbolic agents).
- π 3 papers accepted β ICIC 2026 Γ2 Oral Β· ICML 2026
- 𧬠iGEM 2025 Gold Medal β dry-lab (computational) lead, in-person defence in Paris
- π‘οΈ 1 invention patent filed (neuro-symbolic language-driven travel planning)
- π 10+ academic competition & honour awards (MCM/ICM HM, CUMCM, Lanqiao Cup, Wuxi City Outstanding Student Cadre, Jiangnan Honor Student β¦)
π I am actively looking for research-assistant, summer-internship, and graduate-study opportunities in Computer Vision, AI for Science, and LLM reasoning.
Medical Image Segmentation & Vision-Language Fusion β fine-structure segmentation under weak
semantic prompts, Mamba/SSM, alignment-adaptive fusion
AI4Science β protein & small-molecule design, data analysis & visualisation
AIGC Detection & Visual Robustness β AI-generated image detection, OOD generalisation,
CLIP representations, robustness-asymmetry analysis
1 first author Β·
*co-first author
| # | Paper | Venue | Role | Status |
|---|---|---|---|---|
| 1 | FWMamba-UNet: Frequency-Wavelet Enhanced Mamba UNet for Medical Image Segmentation | ICIC 2026 Oral (CCF-C) | 1st author | β Accepted |
| 2 | Extract Then Compile: Reliable Neuro-Symbolic Planning for Large Language Models | ICIC 2026 Oral (CCF-C) | 1st author | β Accepted |
| 3 | RA-Det: Towards Universal Detection of AI-Generated Images via Robustness Asymmetry | ICML 2026 (CCF-A) | Co-first | β Accepted |
π Full publication list & details on my academic homepage.
|
ICIC 2026 Oral Β· 1st Author Frequency-domain + wavelet-transform branches augmenting a Mamba state-space UNet, capturing cross-scale boundary cues that pure spatial UNets miss.
|
Dry-Lab Lead Β· Paris De-novo antimicrobial peptide design: ESM-2 / ProtT5 / Ankh + BiGRU-VAE β Latent Diffusion β Transformer decoder, three-stage training (MLE β RL β diffusion). SOTA on multiple metrics; wet-lab-validated variants outperform LL-37.
|
|
ICIC 2026 Oral Β· 1st Author + Patent Extract-Then-Compile: an LLM lifts natural-language constraints into a symbolic program, solved reliably by classical search. Filed national invention patent.
|
ICML 2026 Β· Co-First Author Universal detection of AI-generated images via robustness asymmetry β exploiting how real vs. generated images respond differently to perturbations.
|
Deep Learning & LLMs
Architectures & Techniques
Vision & AI4Science
Languages & Tools
