Ze Rong — Research Portfolio

🎓 Undergraduate Student @ Nantong University · Medical Imaging / Multimodal Sports Analytics / Non-invasive BCI

Research Interests

Low-quality enhancement & restoration
Denoise/LLIE for reliable clinical analysis.
🧭
Robust medical image segmentation
Structure-aware masks under domain shifts.
Frequency-domain learning
Signal-aware features for robust med imaging.
🏟️
Tactical analytics & pitch control
Spatiotemporal models for team sports.
🧠
Semantic decoding (fMRI/MEG)
Non-invasive brain→language mappings.
🩺
Vision-Language for medicine
Evidence-grounded clinical reasoning.

🚀 Now / Open To

  • Now: Focusing on non-invasive BCI with fMRI/MEG for semantic decoding and reconstruction, including explorations of inter-regional interactions in the brain.
  • Open to: RA/internship roles and collaborations in aligned areas—feel free to reach out.
  • Seeking: PhD/MPhil opportunities in medical AI / multimodal reasoning (2026 intake).

🔥 News

  • [2025-09] FIRM accepted by ACML 2025.
  • [2025-08] CRAIC 2025 — National Second Prize.
  • [2025-07] FaRMamba accepted by ICONIP 2025.

Featured Projects

FaRMamba — Frequency-aware Mamba for Medical Segmentation
MSFM + reconstruction-aided learning; evaluated on CAMUS / Kvasir-Seg / MICE-cochlea.
Learn more →
FIRM — Fusion-Injected Residual Memory for VI-ReID
VSPF aligning visible↔infrared features; boosts cross-modality matching on common CNN/Transformer backbones.
Learn more →
Ophthalmic Disease Diagnosis Platform
End-to-end eye-disease pipeline: enhance low-quality images → segment key anatomy (FaRMamba) → classify diseases → export concise reports.
Learn more →