no code implementations • 9 Jun 2025 • William Ljungbergh, Bernardo Taveira, Wenzhao Zheng, Adam Tonderski, Chensheng Peng, Fredrik Kahl, Christoffer Petersson, Michael Felsberg, Kurt Keutzer, Masayoshi Tomizuka, Wei Zhan
This is achieved by training R3D2 on a novel dataset: 3DGS object assets are generated from in-the-wild AD data using an image-conditioned 3D generative model, and then synthetically placed into neural rendering-based virtual environments, allowing R3D2 to learn realistic integration.
no code implementations • 5 Jun 2025 • Jinhyung Park, Yihan Hu, Chensheng Peng, Wenzhao Zheng, Kris Kitani, Wei Zhan
Despite the demonstrated efficiency and performance of sparse query-based representations for perception, state-of-the-art 3D occupancy prediction methods still rely on voxel-based or dense Gaussian-based 3D representations.
no code implementations • 1 Dec 2024 • Chensheng Peng, Ido Sobol, Masayoshi Tomizuka, Kurt Keutzer, Chenfeng Xu, Or Litany
Additionally, our method is flexible, as it can learn from various 3D Gaussian Splat (3DGS) teachers with minimal adaptation; we demonstrate this by surpassing the performance of two different deterministic models as teachers, highlighting the potential generalizability of our framework.
1 code implementation • CVPR 2025 • Chensheng Peng, Chengwei Zhang, Yixiao Wang, Chenfeng Xu, Yichen Xie, Wenzhao Zheng, Kurt Keutzer, Masayoshi Tomizuka, Wei Zhan
We present DeSiRe-GS, a self-supervised gaussian splatting representation, enabling effective static-dynamic decomposition and high-fidelity surface reconstruction in complex driving scenarios.
1 code implementation • 2 Nov 2024 • Yichen Xie, Chenfeng Xu, Chensheng Peng, Shuqi Zhao, Nhat Ho, Alexander T. Pham, Mingyu Ding, Masayoshi Tomizuka, Wei Zhan
Recent advancements have exploited diffusion models for the synthesis of either LiDAR point clouds or camera image data in driving scenarios.
no code implementations • CVPR 2025 • Chongjian Ge, Chenfeng Xu, Yuanfeng Ji, Chensheng Peng, Masayoshi Tomizuka, Ping Luo, Mingyu Ding, Varun Jampani, Wei Zhan
To achieve this goal, two core designs are proposed: (1) 3D Gaussians Initialization with 2D compositionality: We transfer the well-established 2D compositionality to initialize the Gaussian parameters on an entity-by-entity basis, ensuring both consistent 3D priors for each entity and reasonable interactions among multiple entities; (2) Dynamic Optimization: We propose a dynamic strategy to optimize 3D Gaussians using Score Distillation Sampling (SDS) loss.
1 code implementation • 27 Mar 2024 • Jiuming Liu, Dong Zhuo, Zhiheng Feng, Siting Zhu, Chensheng Peng, Zhe Liu, Hesheng Wang
Image pixels are pre-organized as pseudo points for image-to-point structure alignment.
1 code implementation • 23 Mar 2024 • Chensheng Peng, Zhaoyu Zeng, Jinling Gao, Jundong Zhou, Masayoshi Tomizuka, Xinbing Wang, Chenghu Zhou, Nanyang Ye
Multiple object tracking is a critical task in autonomous driving.
no code implementations • 12 Mar 2024 • Chensheng Peng, Chenfeng Xu, Yue Wang, Mingyu Ding, Heng Yang, Masayoshi Tomizuka, Kurt Keutzer, Marco Pavone, Wei Zhan
In this paper, we reimagine volumetric representations through the lens of quadrics.
1 code implementation • ICCV 2023 • Chensheng Peng, Guangming Wang, Xian Wan Lo, Xinrui Wu, Chenfeng Xu, Masayoshi Tomizuka, Wei Zhan, Hesheng Wang
Previous methods rarely predict scene flow from the entire point clouds of the scene with one-time inference due to the memory inefficiency and heavy overhead from distance calculation and sorting involved in commonly used farthest point sampling, KNN, and ball query algorithms for local feature aggregation.
no code implementations • 30 Mar 2022 • Guangming Wang, Chensheng Peng, Jinpeng Zhang, Hesheng Wang
Specifically, through multi-scale interactive query and fusion between pixel-level and point-level features, our method, can obtain more distinguishing features to improve the performance of multiple object tracking.