Zhuopeng Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Zhuopeng Li has a total of 11 co-authors. Here are the top co-authors by number of shared publications:
- Jianke Zhu
Zhejiang University
4 shared publications Chenming Wu
Baidu Research
1 shared publication- Chenming Wu
Baidu, Beijing, China
1 shared publication - Huasen He
Department of Automation, University of Science and Technology of China, Hefei, China
1 shared publication Liangjun Zhang
Baidu Research
1 shared publication- Liangjun Zhang
Robotics and Autonomous Driving Laboratory (RAL), Baidu Research, Beijing, China
1 shared publication
Latest Publications
- 26 November 2024
- Published by Springer Nature
- p. 19-36
HO-Gaussian: Hybrid Optimization of 3D Gaussian Splatting for Urban Scenes
- 26 November 2024
- Published by Springer Nature
- p. 19-36
HO-Gaussian: Hybrid Optimization of 3D Gaussian Splatting for Urban Scenes
- 12 November 2024
- Published by Springer Nature
- p. 76-88
Space-View Decoupled 3D Gaussians for Novel-View Synthesis of Mirror Reflections
- 12 November 2024
- Published by Springer Nature
- p. 76-88
Space-View Decoupled 3D Gaussians for Novel-View Synthesis of Mirror Reflections
- 24 June 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Automation Science and Engineering
- Vol. 22 (15455955) , 4394-4407
DGNR: Density-Guided Neural Point Rendering of Large Driving Scenes
Despite the recent success of Neural Radiance Field (NeRF), it is still challenging to render large-scale driving scenes with long trajectories, particularly when the rendering quality and efficiency are in high demand. Existing methods for such scenes usually involve with spatial warping, geometric supervision from zero-shot normal or depth estimation, or scene division strategies, where the synthesized views are often blurry or fail to meet the requirement of efficient rendering. To address the above challenges, this paper presents a novel framework that learns a density space from the scenes to guide the construction of a point-based renderer, dubbed as DGNR (Density-Guided Neural Rendering) . In DGNR, geometric priors are no longer needed, which can be intrinsically learned from the density space through volumetric rendering. Specifically, we make use of a differentiable renderer to synthesize images from the neural density features obtained from the learned density space. A density-based fusion module and geometric regularization are proposed to optimize the density space. By conducting experiments on a widely used autonomous driving dataset, we have validated the effectiveness of DGNR in synthesizing photorealistic driving scenes and achieving real-time capable rendering. Our project page is available at https://github.com/JOP-Lee/DGNR-Rendering. Note to Practitioners —While Neural Radiance Field (NeRF) has been gaining attraction, it is still challenging to create highly detailed, efficient renderings of large driving scenes. Current methods often resort to spatial warping, geometric guidance from tools like zero-shot normal or depth estimates, or dividing the scene into smaller parts. Unfortunately, these techniques can result in blurred images or fail to meet efficiency needs. To solve these challenges, we introduce a learned density space to build a point-based renderer, termed Density-Guided Neural Rendering (DGNR). With DGNR, we no longer need geometric priors because the density space can inherently learn them through volume rendering. Specifically, we use a flexible renderer to create images from the neural density features derived from the learned density space. We have also proposed a density-based fusion module and geometric regularization to optimize the density space. We evaluated DGNR on a popular autonomous driving dataset and found it to be effective in creating realistic driving scenes and capable of real-time rendering. Project page: https://github.com/JOP-Lee/DGNR-Rendering.
- 24 June 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Automation Science and Engineering
- Vol. 22 (15455955) , 4394-4407
DGNR: Density-Guided Neural Point Rendering of Large Driving Scenes
Despite the recent success of Neural Radiance Field (NeRF), it is still challenging to render large-scale driving scenes with long trajectories, particularly when the rendering quality and efficiency are in high demand. Existing methods for such scenes usually involve with spatial warping, geometric supervision from zero-shot normal or depth estimation, or scene division strategies, where the synthesized views are often blurry or fail to meet the requirement of efficient rendering. To address the above challenges, this paper presents a novel framework that learns a density space from the scenes to guide the construction of a point-based renderer, dubbed as DGNR (Density-Guided Neural Rendering) . In DGNR, geometric priors are no longer needed, which can be intrinsically learned from the density space through volumetric rendering. Specifically, we make use of a differentiable renderer to synthesize images from the neural density features obtained from the learned density space. A density-based fusion module and geometric regularization are proposed to optimize the density space. By conducting experiments on a widely used autonomous driving dataset, we have validated the effectiveness of DGNR in synthesizing photorealistic driving scenes and achieving real-time capable rendering. Our project page is available at https://github.com/JOP-Lee/DGNR-Rendering. Note to Practitioners —While Neural Radiance Field (NeRF) has been gaining attraction, it is still challenging to create highly detailed, efficient renderings of large driving scenes. Current methods often resort to spatial warping, geometric guidance from tools like zero-shot normal or depth estimates, or dividing the scene into smaller parts. Unfortunately, these techniques can result in blurred images or fail to meet efficiency needs. To solve these challenges, we introduce a learned density space to build a point-based renderer, termed Density-Guided Neural Rendering (DGNR). With DGNR, we no longer need geometric priors because the density space can inherently learn them through volume rendering. Specifically, we use a flexible renderer to create images from the neural density features derived from the learned density space. We have also proposed a density-based fusion module and geometric regularization to optimize the density space. We evaluated DGNR on a popular autonomous driving dataset and found it to be effective in creating realistic driving scenes and capable of real-time rendering. Project page: https://github.com/JOP-Lee/DGNR-Rendering.
- 14 August 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Vehicles
- Vol. 9 (1) , 2740-2752
Point-Based Neural Scene Rendering for Street Views
While previous neural rendering methods have shown promising results in novel view synthesis, they still face difficulties when dealing with challenging street scenes. For instance, certain objects or content may only appear in a limited number of views, which can result in blurry synthesized images with significant artifacts. To address this critical issue, we propose a novel point-based neural rendering framework that facilitates a photo-realistic street simulation environment. Specifically, we introduce the Neural Street Scene Renderer (NSSR) to handle holes resulting from sparse point clouds. We also present a parallel attention module (PAM) and a multi-scale feature fusion and selection module (FFSM) to recalibrate features. Additionally, we use a point cloud filter and Gaussian noise points to remove outliers or floating points in the air and repair sparse point clouds. Our experimental results on four datasets demonstrate the effectiveness of our proposed approach in synthesizing photo-realistic street scenes.
- 14 August 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Vehicles
- Vol. 9 (1) , 2740-2752
Point-Based Neural Scene Rendering for Street Views
While previous neural rendering methods have shown promising results in novel view synthesis, they still face difficulties when dealing with challenging street scenes. For instance, certain objects or content may only appear in a limited number of views, which can result in blurry synthesized images with significant artifacts. To address this critical issue, we propose a novel point-based neural rendering framework that facilitates a photo-realistic street simulation environment. Specifically, we introduce the Neural Street Scene Renderer (NSSR) to handle holes resulting from sparse point clouds. We also present a parallel attention module (PAM) and a multi-scale feature fusion and selection module (FFSM) to recalibrate features. Additionally, we use a point cloud filter and Gaussian noise points to remove outliers or floating points in the air and repair sparse point clouds. Our experimental results on four datasets demonstrate the effectiveness of our proposed approach in synthesizing photo-realistic street scenes.
- Open Access
- 26 June 2023
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 37 (2) , 1522-1529
READ: Large-Scale Neural Scene Rendering for Autonomous Driving
With the development of advanced driver assistance systems~(ADAS) and autonomous vehicles, conducting experiments in various scenarios becomes an urgent need. Although having been capable of synthesizing photo-realistic street scenes, conventional image-to-image translation methods cannot produce coherent scenes due to the lack of 3D information. In this paper, a large-scale neural rendering method is proposed to synthesize the autonomous driving scene~(READ), which makes it possible to generate large-scale driving scenes in real time on a PC through a variety of sampling schemes. In order to effectively represent driving scenarios, we propose an ω-net rendering network to learn neural descriptors from sparse point clouds. Our model can not only synthesize photo-realistic driving scenes but also stitch and edit them. The promising experimental results show that our model performs well in large-scale driving scenarios.
- 26 June 2023
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 37 (2) , 1522-1529
Open AccessREAD: Large-Scale Neural Scene Rendering for Autonomous Driving
With the development of advanced driver assistance systems~(ADAS) and autonomous vehicles, conducting experiments in various scenarios becomes an urgent need. Although having been capable of synthesizing photo-realistic street scenes, conventional image-to-image translation methods cannot produce coherent scenes due to the lack of 3D information. In this paper, a large-scale neural rendering method is proposed to synthesize the autonomous driving scene~(READ), which makes it possible to generate large-scale driving scenes in real time on a PC through a variety of sampling schemes. In order to effectively represent driving scenarios, we propose an ω-net rendering network to learn neural descriptors from sparse point clouds. Our model can not only synthesize photo-realistic driving scenes but also stitch and edit them. The promising experimental results show that our model performs well in large-scale driving scenarios.