Hi there! I am Yuan Wu (吴渊), a Ph.D. student at PCALab, School of Computer Science and Engineering, Nanjing University of Science and Technology. I am fortunate to be supervised by Prof. Jian Yang and co-supervised by Dr. Zhiqiang Yan. My research interests lie in autonomous driving perception, particularly 3D occupancy prediction. Feel free to reach out with any questions or suggestions! 😊
📝 Publications
Height-Guided Projection Reparameterization for Camera-LiDAR Occupancy
Most previous methods rely on a fixed projection space, where 3D reference points are uniformly sampled along pillars. However, such sampling struggles to capture the sparsity and height variations of real-world scenes, leading to ambiguous correspondences and unreliable feature aggregation. To address these challenges, we propose HiPR, a camera-LiDAR occupancy framework with Height-Guided Projection Reparameterization.
Zero-Shot Depth Completion with Vision Language Model
This paper introduces the first VLM-based depth completion framework. We propose a sparse depth injection mechanism that extends VLM's capability toward 3D perception through three key aspects: visual tokenization, textual prompt, and textual supervision.
SpatioTemporal Difference Network for Video Depth Super-Resolution Oral
We propose STDNet, a novel framework for video depth super-resolution. STDNet introduces spatial and temporal difference mechanisms to mitigate long-tailed effects in video depth super-resolution. This design enables precise depth calibration and motion compensation, leading to state-of-the-art performance.
See through the Dark: Learning Illumination-affined Representations for Nighttime Occupancy Prediction
Existing vision-based methods perform well on daytime benchmarks but struggle in nighttime scenarios due to limited visibility and challenging lighting conditions. We introduce LIAR, a novel framework that learns illumination-affined representations for nighttime occupacy prediction.
Deep Height Decoupling for Precise Vision-based 3D Occupancy Prediction
For the first time, we introduce the explicit height prior into the vision-based 3D occupancy predition task. Owing to the novel deep height decoupling and sampling stratagy, our model achieves state-of-the-art performance even with minimal input cost.
🎖 Honors and Awards
- 2024.06: Outstanding Graduates of Nanjing University of Science and Technology
📖 Educations
- 2024.09 - present: Ph.D. student, School of Computer Science and Engineering, Nanjing University of Science and Technology
- 2020.09 - 2024.06: B.Eng., School of Intelligent Manufacturing, Nanjing University of Science and Technology