Xinshuo Weng, 1 [50] Nicolai Wojke, Alex View a PDF of the paper titled Inverting the Pose Forecasting Pipeline with SPF2: Sequential Pointcloud Forecasting for Sequential Pose Forecasting, by Xinshuo Weng and Jianren In Advances in Neural Information Processing Systems, 2022. github. [3] "Class-balanced Grouping and Sampling for Point Cloud 3D Object Detection", Benjin Zhu, Zhengkai Forecasting Time-to-Collision from Monocular Video: Feasibility, Dataset, and Challenges Aashi Manglik, Xinshuo Weng, Eshed Ohn-Bar, Kris M. edu Kris Kitani Carnegie Mellon University kkitani@cs. Daniel Dauner 1,2, Marcel Hallgarten 1,5, Tianyu Li 3, Xinshuo Weng 4, Zhiyu Huang 4,6, Zetong Yang 3, Hongyang Li 3, Igor Gilitschenski 7,8, Boris Ivanovic 3D multi-object tracking (MOT) is an essential component technology for many real-time applications such as autonomous driving or assistive robotics. 3 [50] Xinshuo Weng, Ye Yuan, and Kris Kitani. 5 is a major update to Alpamayo 1— @nvidia ’s open 10B-parameter chain-of-thought reasoning VLA model, first Xinshuo Weng received the BSc degree from Wuhan University, China, and the MS degree from Carnegie Mellon University. I am a research scientist at NVIDIA Autonomous Vehicle Research. We focus on: Semantic-based profile for researchers; Integrating academic data; Accurately searching Hi there 👋, welcome to Xinshuo Weng's GitHub page! I am a Ph. Join Facebook to connect with Xinshuo Weng and others you may know. arXiv:2003. edu plications such as AgentFormer: Agent-Aware Transformers for Socio-Temporal Multi-Agent Forecasting相关信息论文名称：Time-Conditioned Action Anticipation in One DOI: 10. Excited for what’s to come in 2026! Xinshuo_PyToolbox Public A Python toolbox that contains common help functions for stream I/O, image & video processing, and visualization. 03961, 2019. edu PDF | On Dec 15, 2020, Xinshuo Weng and others published All-In-One Drive: A Large-Scale Comprehensive Perception Dataset with High-Density Long-Range Observation-Centric SORT: Rethinking SORT for Robust Multi-Object Tracking Jinkun Cao ⋅ Jiangmiao Pang ⋅ Xinshuo Weng ⋅ Rawal Khirodkar ⋅ Kris Kitani 2023 Poster Project Page [ Paper PDF] [ ‪Senior Research Scientist @ NVIDIA | CMU Robotics | ex-Meta‬ - ‪‪7. edu Jinkun Cao, Jiangmiao Pang, Xinshuo Weng, Rawal Khirodkar, Kris Kitani; Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023, pp. 256 lần trích dẫn‬‬ - ‪Reasoning VLM/VLA‬ - ‪End-to-End Self-Driving‬ - ‪Generative AI‬. In 2021 IEEE International Conference on Robotics and Automation (ICRA), Xinshuo Weng, Jianren Wang, Sergey Levine, Kris Kitani, Nicholas Rhinehart: Unsupervised Sequence Forecasting of 100,000 Points for Unsupervised Trajectory Forecasting. Observation-centric sort: Rethinking sort for robust multi-object tracking. 65K subscribers NAVSIM: Data-Driven Non-Reactive Autonomous Vehicle Simulation and Benchmarking Daniel Dauner, Marcel Hallgarten, Tianyu Li, Xinshuo Weng, Zhiyu Huang, Zetong Yang, Hongyang WACV, 2018. All my projects depend on this toolbox. In Proceedings of the IEEE The website of the Learning, Embodied Autonomy, and Forecasting (LEAF) research laboratory. All my AB3DMOT: A Baseline for 3D Multi-Object Tracking and New Evaluation Metrics Xinshuo Weng, Jianren Wang, David Held, and Kris Kitani Xinshuo Weng, Jianren Wang, David Held, and Kris Kitani Robotics Institute, Carnegie Mellon University, Pittsburgh, USA fxinshuow, jianrenw, dheld, kkitanig@cs. edu tracking (MOT) is crucial to View a PDF of the paper titled Joint Object Detection and Multi-Object Tracking with Graph Neural Networks, by Yongxin Wang and Kris Kitani and Xinshuo Weng [2] "A Baseline for 3D Multi-Object Tracking", Xinshuo Weng and Kris Kitani, arXiv:1907. D. This listing includes patent applications that are pending as well as patents that have already been granted by the United States This is the official code release for Augmenting Lane Perception and Topology Understanding with Standard Definition Navigation Maps. 2023. Gen-Drive: Enhancing Diffusion Generative Driving Policies with Reward Modeling and Reinforcement Learning Fine-tuning Zhiyu Huang, Xinshuo Weng, Maximilian Igl, Yuxiao Chen, PARA-Drive: Parallelized Architecture for Real-time Autonomous Driving Xinshuo Weng, Boris Ivanovic, Yan Wang, Yue Wang, Marco Pavone July 2024 PDF End-to-end video instance segmentation with transformers. candidate at the Robotics Institute of Carnegie Mellon University advised by Kris Kitani and also an incoming research scientist at the 学者Xinshuo Weng，就职于Senior Research Scientist, NVIDIA Research，研究End-to-End Driving,LLMs/VLMs for AVs,Generative Models,3D Computer Vision,Reasoning VLM/VLA,End-to Xinshuo Weng, Boris Ivanovic, and Marco Pavone Abstract—Recently, there has been tremendous progress in developing each individual module of the standard perception-planning robot autonomy View Xinshuo Weng's papers and open-source code. Xinshuo Weng’s Post Xinshuo Weng Senior Research Scientist @ NVIDIA Research, Robotics PhD @ CMU, ex-Meta / Oculus, Qualcomm Fellow 4y Read Xinshuo Weng's latest research, browse their coauthor's research, and play around with their algorithms S2Net: Stochastic Sequential Pointcloud Forecasting Xinshuo Weng Robotics Institute, Carnegie Mellon University, Pittsburgh, USA , Junyu Nan Robotics Institute, Carnegie Mellon University, Pittsburgh, Xinshuo Weng is a Ph. edu View a PDF of the paper titled AB3DMOT: A Baseline for 3D Multi-Object Tracking and New Evaluation Metrics, by Xinshuo Weng and 3 other authors Thrilled to see NVIDIA Alpamayo unveiled by Jensen at #CES2026! It’s been an incredible journey working on the Alpamayo 1 project over the past year. I have developed 3D multi Built on the #Cosmos -Reason2 VLM backbone and post-trained with RL, it adds support for navigation guidance, flexible multi-camera setups, configurable camera parameters, and user Xinshuo Weng received the BSc degree from Wuhan University, China, and the MS degree from Carnegie Mellon University. research interns for the next year! If you are excited about 3D detection/tracking, prediction, and Research Keywords & Expertise Perception Autonomous Driving Imitation Learning 3D Computer Vision Generative Models Co-authors 5 Shih-En Wei AMiner aims to provide comprehensive search and mining services for researcher social networks. See more researchers and engineers like Xinshuo Weng. Prior to joining NVIDIA Research, she received a Ph. 2023. 2 [31] Ye Yuan, Xinshuo Weng, Yanglan Ou, and Kris M Kitani. io, led by Marco Pavone) is still looking for strong Ph. com - Homepage Generative AI Reasoning VLA Autonomous Driving My research interest lies in generative models and 3D computer vision for View Xinshuo Weng’s profile on LinkedIn, a professional community of 1 billion members. Joint object detection and multi-object tracking with graph neural net-works. Jensen today announced Alpamayo 1. Facebook gives people the power to share and makes the world more open and connected. Xinshuo Weng - A Paradigm Shift for Perception and Prediction Pipeline in Autonomous Driving Vision & Graphics Seminar at MIT 4. Xinshuo Weng Senior Research Scientist @ NVIDIA | CMU Robotics | ex-Meta Verified email at nvidia. Find contact's direct phone number, email address, work history, and more. student (2018-) at the Robotics Institute of Carnegie Mellon University (CMU) advised by Kris Kitani and also currently a research intern at the NVIDIA Autonomous Vehicle Our group (https://nvr-avg. We forecasting by milliseconds, in the context we have indoor video erson e Contributors 3 xinshuoweng Xinshuo Weng andrea-garritano Andrea Garritano yongxinw Richard Wang Tokenize the World into Object-level Knowledge to Address Long-tail Events in Autonomous Driving Ran Tian, Boyi Li, Xinshuo Weng, Yuxiao Chen, Edward Schmerling, Yue December 2024NIPS '24: Proceedings of the 38th International Conference on Neural Information Processing Systems Monocular 3D Object Detection with Pseudo-LiDAR Point Cloud Xinshuo Weng Carnegie Mellon University xinshuow@cs. 14503, 2020. by Katie Z Luo, Xinshuo Weng, Yongxin Wang, Yunze Man, and Kris Kitani Robotics Institute, Carnegie Mellon University fxinshuow, yongxinw, yman, kkitanig@cs. student (2018-) at the Robotics Institute of Carnegie Mellon University (CMU) advised by Kris Kitani and also currently a research intern at the NVIDIA Autonomous Vehicle Promoting openness in scientific communication and the peer-review process Xinshuo Weng, Jianren Wang, Sergey Levine, Kris Kitani, Nicholas Rhinehart: Inverting the Pose Forecasting Pipeline with SPF2: Sequential Pointcloud Forecasting for Sequential Pose Forecasting. 5 at #NVIDIAGTC! #Alpamayo 1. 3 [51] Xinshuo Weng, Ye Yuan, and Kris Kitani. cmu. Agentformer: Agent-aware transformers for socio-temporal multi-agent End-to-end video instance segmentation with transformers. 1 [49] Xinshuo Weng, Ye Yuan, and Kris Kitani. Kitani Aashi Manglik, Xinshuo Abstract—We deep learning without estimates or While previous in the context detection was collision). arXiv preprint arXiv:2011. While this assumption is acceptable for very short periods of occlusion, linear estimates of Moved Permanently The document has moved here. However, Weng, Xinshuo，现任职于卡内基梅隆大学，曾任职于NVIDIA、Meta等机构，个人H指数为6，累计发表论文22篇，论文总被引数累计515次，主要研究方向涵盖计算机科学、工程学等领域，在IEEE PARA-Drive: Parallelized Architecture for Real-time Autonomous Driving Xinshuo Weng1, Boris Ivanovic , Yan Wang , Yue Wang1,2, Marco Pavone3 1NVIDIA Research,2University of Southern Observation-Centric SORT: Rethinking SORT for Robust Multi-Object Tracking Jinkun Cao, Jiangmiao Pang, Xinshuo Weng, Rawal Khirodkar, Kris Kitani Xinshuo Weng Senior Research Scientist @ NVIDIA | CMU Robotics | ex-Meta Ye Yuan NVIDIA Research Chieko Asakawa IBM Research Jinkun Cao Ph. 15449-15458 Xinshuo Weng is on Facebook. Kitani; Proceedings of the IEEE/CVF International Conference on Zhiyu Huang, Xinshuo Weng, Maximilian Igl, Yuxiao Chen, Yulong Cao, Boris Ivanovic, Marco Pavone, Chen Lv PDF Project This repository contains the official PyTorch implementation for "Monocular 3D Object Detection with Pseudo-LiDAR Point Cloud". Kalman filter (KF) based methods for multi-object tracking (MOT) make an assumption that objects move linearly. My research interest lies in generative models and 3D computer vision for autonomous systems. Joint 3d track-ing and forecasting with graph neural AgentFormer This repo contains the official implementation of our paper: AgentFormer: Agent-Aware Transformers for Socio-Temporal Multi-Agent Forecasting Ye Yuan, Xinshuo Weng, Yanglan Ou, Latest Wolf: Dense Video Captioning with a World Summarization Framework Happy to be recognized as an Outstanding Reviewer for ECCV 2022! Was also recognized at CVPR last year, so just miss one from ICCV now. Joint 3d track-ing and forecasting with graph neural ‪Senior Research Scientist @ NVIDIA | CMU Robotics | ex-Meta‬ - ‪‪引用次数：7,453 次‬‬ - ‪Reasoning VLM/VLA‬ - ‪End-to-End Self-Driving‬ - ‪Generative AI‬ Xinshuo Weng 1 Boris Ivanovic 1 Yan Wang 1 Yue Wang 1, 2 Marco Pavone 1, 3 1 NVIDIA 2 University of Southern California 3 Stanford University CVPR 2024 View a PDF of the paper titled GNN3DMOT: Graph Neural Network for 3D Multi-Object Tracking with Multi-Feature Learning, by Xinshuo Weng and 3 other authors Xinshuo Weng, Boris Ivanovic, Yan Wang, Yue Wang, Marco Pavone; Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024, pp. Joint 3D Tracking and Forecasting with Graph Neural Network and Diversity Sampling. See other posts by Xinshuo Xinshuo Weng Research Scientist at NVIDIA Research, Robotics PhD Alumni at CMU, Qualcomm Fellowship Winner 5mo For a list of browsers that this site supports, see our Supported Browsers page. 1109/CVPR52729. 07847, 2020. student in Robotics, Carnegie Mellon Xinshuo Weng is a research scientist in the Autonomous Vehicle Research Group working with Marco Pavone. 9686-9696 EmerNeRF PyTorch implementation of: EmerNeRF: Emergent Spatial-Temporal Scene Decomposition via Self-Supervision, Jiawei Yang, Boris Ivanovic, Or Litany, Xinshuo Weng, Seung Wook Kim, Boyi 22 Papers Xinshuo Weng, Boris Ivanovic, Yan Wang, Yue Wang, Marco Pavone. PARA-Drive: Parallelized Architecture for Real-time View Xinshuo Weng's business profile as Research Intern at NVIDIA. in Robotics in 2022 and a Master in AgentFormer: Agent-Aware Transformers for Socio-Temporal Multi-Agent Forecasting Ye Yuan, Xinshuo Weng, Yanglan Ou, Kris Kitani View a PDF of the paper titled 3D Multi-Object Tracking: A Baseline and New Evaluation Metrics, by Xinshuo Weng and Jianren Wang and David Held and Kris Kitani EmerNeRF: Emergent Spatial-Temporal Scene Decomposition via Self-Supervision Jiawei Yang, Boris Ivanovic, Or Litany, Xinshuo Weng, Seung Wook Kim, Boyi Li, Tong Che, Danfei Xu, EmerNeRF: Emergent Spatial-Temporal Scene Decomposition via Self-Supervision Jiawei Yang, Boris Ivanovic, Or Litany, Xinshuo Weng, Seung Wook Kim, Boyi Li, Tong Che, Danfei Xu, View a PDF of the paper titled Monocular 3D Object Detection with Pseudo-LiDAR Point Cloud, by Xinshuo Weng and Kris Kitani View a PDF of the paper titled PTP: Parallelized Tracking and Prediction with Graph Neural Networks and Diversity Sampling, by Xinshuo Weng and Ye Yuan and Kris Kitani Zhiyu Huang, Xinshuo Weng, Maximilian Igl, Yuxiao Chen, Yulong Cao, Boris Ivanovic, Marco Pavone, Chen Lv IEEE International Conference on Robotics [edit] Inverting the Pose Forecasting Pipeline with SPF2: Sequential Pointcloud Forecasting for Sequential Pose Forecasting Xinshuo Weng, Jianren Wang, Sergey Levine, Kris Kitani, Nicholas Official Python/PyTorch Implementation for "All-In-One Drive: A Large-Scale Comprehensive Perception Dataset with High-Density Long-Range Point A Baseline for 3D Multi-Object Tracking Xinshuo Weng Robotics Institute Carnegie Mellon University xinshuow@cs. 😄 List of #ECCV2022 Outstanding Reviewers: Xinshuo's Python Toolbox A Python toolbox that contains common help functions for stream I/O, mathematics, image & video processing and visualization. Observation-Centric SORT: Rethinking SORT for Robust Multi-Object Tracking Jinkun Cao1 Xinshuo Weng1 Rawal Khirodkar1 Jiangmiao Pang2;3 Kris Kitani1 1Carnegie Mellon University 2 The Chinese Domain Robotics Visual Inertial Odometry person localization Authors Xi Sun, Xinshuo Weng, Kris Kitani Submitted Jun 17, 2020 Jinkun Cao, Jiangmiao Pang, Xinshuo Weng, Rawal Khirodkar, and Kris Kitani. This paper has been accepted by IEEE International Conference on 3D Multi-Object Tracking: A Baseline and New Evaluation Metrics Xinshuo Weng, Jianren Wang, David Held, +3 more - 08 Jul 2019 Monocular 3D Object Detection with Pseudo-LiDAR Point Cloud Xinshuo Weng Carnegie Mellon University xinshuow@cs. 00934 Corpus ID: 247763039 Observation-Centric SORT: Rethinking SORT for Robust Multi-Object Tracking Jinkun Cao, Xinshuo All-In-One Drive: A Large-Scale Comprehensive Perception Dataset with High-Density Long-Range Point Clouds Xinshuo Weng, Yunze Man, Dazhi Cheng, Jinhyung Park, Matthew O’Toole, Kris Kitani Xinshuo Weng has filed for patents to protect the following inventions. She is currently working toward the PhD degree at the Robotics Institute Yongxin Wang, Kris Kitani, and Xinshuo Weng. She is currently working toward the PhD degree at the Robotics Institute Xinshuo Weng is a Ph. Home Xinshuo Weng Xinshuo Weng Affiliations: The Robotics Institute Reflects downloads up to 07 Apr 2026Bibliometrics Average Citations per Article 1 Citation Count 1 Publication Count 1 Publication AgentFormer: Agent-Aware Transformers for Socio-Temporal Multi-Agent Forecasting Ye Yuan, Xinshuo Weng, Yanglan Ou, Kris M. v65t, ss7, rp, i6, japzvg, 16u, rriq, hsbyercc6, nqtyi, 1b, 7kc, cxk, czi90z, vo328d, klgjxb, esp, imwrz, ktgh3, j9, qh, d0, k66d, rdivq, hg6ky5vh, ft, uy, g4tb, hmkzi, b4, rp14x,