Hi, I’m Zongqi He, currently a final year undergraduate student from Department of Electrical and Electronic Engineering, The Hong Kong Polytechnic University. I am very fortunate to be advised by Prof. Kenneth K. M. Lam.

My research interest includes computer vision, deep learning , low-level vision and 3D reconstruction.

Here is my CV and transcripts.

🔥 News

2025.04: 🎉🎉 Our team has submitted two papers to ACM MM 2025, where I contributed as first author on one and co-first author on the other.
2025.03: 🎉🎉 Our team ranked among the top in three NTIRE 2025 Challenges.
2025.01: 🎉🎉 Our paper on enhancing Sparse input 3D Gaussian splatting for novel view synthesis is submitted to IEEE Transactions on Visualization and Computer Graphics (TVCG).
2025.01: 🎉🎉 Our paper See In Detail: Enhancing Sparse-view 3D Gaussian Splatting with Local Depth and Semantic Regularization has been accepted by International Conference on Acoustics, Speech, and Signal Processin (ICASSP) 2025.
2024.10: 🎉🎉 Our paper MFGan: OCT Image Super-resolution and Enhancement with Blind Degradation and Multi-frame Fusing is accpeted by International Workshop on Advanced Image Technology (IWAIT) 2025.
2024.10: 🎉🎉 Our paper A Multi-Perceptual Learning Network for Retina OCT Image Denoising and Classification is accpeted by Asia-Pacific Signal and Information Processing Association (APSIPA) 2024.

🚧 Ongoing Projects

Gaussian Avatar: Developing a dynamic avatar system using 3D Gaussian Splatting to enable expressive and real-time human rendering.
Novel View Synthesis: Enhancing Sparse-view Gaussian Splatting with Multi-view Consistent Diffusion for 360° Novel View Synthesis from Unposed Viewpoints.
Test Time Scaling: Exploring test-time scaling strategies within cutting-edge text-to-image generation frameworks.

ICASSP 2025

Zongqi He, Zhe Xiao, Kin-Chung Chan, Yushen Zuo, Jun Xiao, Kin-Man Lam.

AI4VA@ECCV 2024

Towards Multi-View Consistent Style Transfer with One-Step Diffusion via Vision Conditioning

Yushen Zuo, Jun Xiao, Kin-Chung Chan, Rongkang Dong, Cuixin Yang, Zongqi He, Hao Xie, Kin-Man Lam

2025.03 NTIRE 2025 Challenge on Night Photography Rendering - 5th place.
2025.03 NTIRE 2025 Challenge on Ambient Light Normalization - 6th place.
2025.03 NTIRE 2025 Challenge on Restore Any Image Model (RAIM) in the Wild - Track 1 - 3rd place.
2024.08 AIM 2024 Challenge on Sparse Neural Rendering - Track 1 - 3 views - 3rd place.
2024.08 AIM 2024 Challenge on Sparse Neural Rendering - Track 2 - 9 views - 3rd place.
2024.08 AIM 2024 Challenge on Efficient Video Super-Resolution for AV1 Compressed Content - 2nd place.
2022-2024 Dean’s Honours List (two years)
2023 HKSAR Government Talent Development Scholarship

UV prediction using GNN
- I am working as a student assistant under Prof. Mengying Li, focusing on UV index prediction around Hong Kong using Graph Neural Networks (GNNs).
Enhancing Sparse-view 3D Gaussian Splatting
- We proposed a novel 3D Gaussian Splatting method, SIDGaussian, for novel view synthesis under extremely sparse input conditions. We designed a semantic regularization and local depth regularization.
  - 📝 This work has been accepted by ICASSP 2025.
- After that, we explored the potential of leveraging diffusion models to enhance sparse-view 3DGS.
  - Our paper has been submitted to ACM MM 2025.
Medical Image Analysis
- Built a novel visual odometry pipeline for colonoscopic navigation, achieving robust depth and pose estimation under illumination shifts.
  - Our paper has been submitted to ACM MM 2025.
Retina OCT Image Denoising and Classification
- Proposed the FD Loss into the GAN architecture, which helps preserve the structural integrity of OCT images during denoising.
  - 📝 This work has been accepted by APSIPA 2024.