no code implementations • 2 May 2024 • Guangming Wang, Lei Pan, Songyou Peng, Shaohui Liu, Chenfeng Xu, Yanzi Miao, Wei Zhan, Masayoshi Tomizuka, Marc Pollefeys, Hesheng Wang
Meticulous 3D environment representations have been a longstanding goal in computer vision and robotics fields.
no code implementations • 15 Apr 2024 • Yujie Yang, Zhilong Zheng, Shengbo Eben Li, Masayoshi Tomizuka, Changliu Liu
We demonstrate our feasibility theory by visualizing different feasible regions under both MPC and RL policies in an emergency braking control task.
1 code implementation • 9 Apr 2024 • Tong Zhao, Lei Yang, Yichen Xie, Mingyu Ding, Masayoshi Tomizuka, Yintao Wei
This paper uniformly proposes two simple yet effective models for road elevation reconstruction in BEV named RoadBEV-mono and RoadBEV-stereo, which estimate road elevation with monocular and stereo images, respectively.
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
This focus results in a significant disconnect between NeRF applications, i. e., novel-view synthesis and the requirements of SLAM.
no code implementations • 10 Mar 2024 • Juanwu Lu, Wei Zhan, Masayoshi Tomizuka, Yeping Hu
For interpretability, the model achieves target-driven motion prediction by estimating the spatial distribution of long-term destinations with a variational mixture of Gaussians.
no code implementations • 9 Mar 2024 • Zhuo Xu, Rui Zhou, Yida Yin, Huidong Gao, Masayoshi Tomizuka, Jiachen Li
Data-driven methods have great advantages in modeling complicated human behavioral dynamics and dealing with many human-robot interaction applications.
no code implementations • 26 Feb 2024 • Dingkun Guo, Yuqi Xiang, Shuqi Zhao, Xinghao Zhu, Masayoshi Tomizuka, Mingyu Ding, Wei Zhan
With these two capabilities, PhyGrasp is able to accurately assess the physical properties of object parts and determine optimal grasping poses.
no code implementations • 23 Feb 2024 • Yichen Xie, Hongge Chen, Gregory P. Meyer, Yong Jae Lee, Eric M. Wolff, Masayoshi Tomizuka, Wei Zhan, Yuning Chai, Xin Huang
Observations from different angles enable the recovery of 3D object states from 2D image inputs if we can identify the same instance in different input frames.
no code implementations • 22 Feb 2024 • Catherine Weaver, Chen Tang, Ce Hao, Kenta Kawamoto, Masayoshi Tomizuka, Wei Zhan
Thus, we propose BeTAIL: Behavior Transformer Adversarial Imitation Learning, which combines a Behavior Transformer (BeT) policy from human demonstrations with online AIL.
1 code implementation • 14 Feb 2024 • Tong Zhao, Mingyu Ding, Wei Zhan, Masayoshi Tomizuka, Yintao Wei
Furthermore, we propose a more rigorous evaluation metric that considers depth-wise relative error, providing comprehensive evaluations for universal stereo matching and depth estimation models.
no code implementations • 31 Dec 2023 • Wei-Jer Chang, Francesco Pittaluga, Masayoshi Tomizuka, Wei Zhan, Manmohan Chandraker
These findings affirm that guided diffusion models provide a robust and versatile foundation for safety-critical, interactive traffic simulation, extending their utility across the broader landscape of autonomous driving.
1 code implementation • 18 Dec 2023 • Zhixuan Liang, Yao Mu, Hengbo Ma, Masayoshi Tomizuka, Mingyu Ding, Ping Luo
Experiments on multi-task robotic manipulation benchmarks like Meta-World and LOReL demonstrate state-of-the-art performance and human-interpretable skill representations from SkillDiffuser.
no code implementations • 17 Dec 2023 • Xinghao Zhu, Devesh K. Jha, Diego Romeres, Lingfeng Sun, Masayoshi Tomizuka, Anoop Cherian
Automating the assembly of objects from their parts is a complex problem with innumerable applications in manufacturing, maintenance, and recycling.
no code implementations • 11 Dec 2023 • Lingfeng Sun, Devesh K. Jha, Chiori Hori, Siddarth Jain, Radu Corcodel, Xinghao Zhu, Masayoshi Tomizuka, Diego Romeres
Designing robotic agents to perform open vocabulary tasks has been the long-standing goal in robotics and AI.
no code implementations • 3 Nov 2023 • Tommaso Benciolini, Chen Tang, Marion Leibold, Catherine Weaver, Masayoshi Tomizuka, Wei Zhan
In the exploration, a MPC collects diverse data by balancing the racing objectives and the exploration criterion; then the GP is re-trained.
no code implementations • 11 Oct 2023 • Yuxin Chen, Chen Tang, Ran Tian, Chenran Li, Jinning Li, Masayoshi Tomizuka, Wei Zhan
We observe that, generally, a more diverse set of co-play agents during training enhances the generalization performance of the ego agent; however, this improvement varies across distinct scenarios and environments.
no code implementations • 11 Oct 2023 • Ran Tian, Chenfeng Xu, Masayoshi Tomizuka, Jitendra Malik, Andrea Bajcsy
When operating in service of people, robots need to optimize rewards aligned with end-user preferences.
no code implementations • 4 Oct 2023 • Mingxiao Huo, Mingyu Ding, Chenfeng Xu, Thomas Tian, Xinghao Zhu, Yao Mu, Lingfeng Sun, Masayoshi Tomizuka, Wei Zhan
We introduce Task Fusion Decoder as a plug-and-play embedding translator that utilizes the underlying relationships among these perceptual skills to guide the representation learning towards encoding meaningful structure for what's important for all perceptual skills, ultimately empowering learning of downstream robotic manipulation tasks.
no code implementations • 4 Oct 2023 • Hao Sha, Yao Mu, YuXuan Jiang, Li Chen, Chenfeng Xu, Ping Luo, Shengbo Eben Li, Masayoshi Tomizuka, Wei Zhan, Mingyu Ding
Existing learning-based autonomous driving (AD) systems face challenges in comprehending high-level information, generalizing to rare events, and providing interpretability.
no code implementations • 3 Oct 2023 • Tong Zhao, Chenfeng Xu, Mingyu Ding, Masayoshi Tomizuka, Wei Zhan, Yintao Wei
This paper addresses the growing demands for safety and comfort in intelligent robot systems, particularly autonomous vehicles, where road conditions play a pivotal role in overall driving performance.
no code implementations • 3 Oct 2023 • Haoyu Zhou, Mingyu Ding, Weikun Peng, Masayoshi Tomizuka, Lin Shao, Chuang Gan
This work introduces a framework harnessing the capabilities of Large Language Models (LLMs) to generate primitive task conditions for generalizable long-horizon manipulations with novel objects and unseen tasks.
1 code implementation • NeurIPS 2023 • Yichen Xie, Mingyu Ding, Masayoshi Tomizuka, Wei Zhan
However, current approaches, represented by active learning methods, typically follow a cumbersome pipeline that iterates the time-consuming model training and batch data selection repeatedly.
no code implementations • 18 Sep 2023 • Jinning Li, Xinyi Liu, Banghua Zhu, Jiantao Jiao, Masayoshi Tomizuka, Chen Tang, Wei Zhan
GOLD distills an offline DT policy into a lightweight policy network through guided online safe RL training, which outperforms both the offline DT policy and online safe RL algorithms.
no code implementations • 18 Sep 2023 • Yiheng Li, Seth Z. Zhao, Chenfeng Xu, Chen Tang, Chenran Li, Mingyu Ding, Masayoshi Tomizuka, Wei Zhan
We propose to augment both HD maps and trajectories and apply pre-training strategies on top of them.
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.
2 code implementations • ICCV 2023 • Chenfeng Xu, Bichen Wu, Ji Hou, Sam Tsai, RuiLong Li, Jialiang Wang, Wei Zhan, Zijian He, Peter Vajda, Kurt Keutzer, Masayoshi Tomizuka
We present NeRF-Det, a novel method for indoor 3D detection with posed RGB images as input.
no code implementations • 29 Jun 2023 • Zitian Chen, Mingyu Ding, Yikang Shen, Wei Zhan, Masayoshi Tomizuka, Erik Learned-Miller, Chuang Gan
We present a model that can perform multiple vision tasks and can be adapted to other downstream tasks efficiently.
no code implementations • NeurIPS 2023 • Chenran Li, Chen Tang, Haruki Nishimura, Jean Mercat, Masayoshi Tomizuka, Wei Zhan
Specifically, we formulate the customization problem as a Markov Decision Process (MDP) with a reward function that combines 1) the inherent reward of the demonstration; and 2) the add-on reward specified by the downstream task.
no code implementations • 14 Jun 2023 • Ce Hao, Catherine Weaver, Chen Tang, Kenta Kawamoto, Masayoshi Tomizuka, Wei Zhan
Hierarchical reinforcement learning (RL) can accelerate long-horizon decision-making by temporally abstracting a policy into multiple levels.
no code implementations • 2 Jun 2023 • Lingfeng Sun, Haichao Zhang, Wei Xu, Masayoshi Tomizuka
In this work, we investigate the potential of improving multi-task training and also leveraging it for transferring in the reinforcement learning setting.
no code implementations • 27 Apr 2023 • Chao Xia, Chenfeng Xu, Patrick Rim, Mingyu Ding, Nanning Zheng, Kurt Keutzer, Masayoshi Tomizuka, Wei Zhan
Current LiDAR odometry, mapping and localization methods leverage point-wise representations of 3D scenes and achieve high accuracy in autonomous driving tasks.
1 code implementation • ICCV 2023 • Yichen Xie, Chenfeng Xu, Marie-Julie Rakotosaona, Patrick Rim, Federico Tombari, Kurt Keutzer, Masayoshi Tomizuka, Wei Zhan
However, given that objects occupy only a small part of a scene, finding dense candidates and generating dense representations is noisy and inefficient.
1 code implementation • CVPR 2023 • Yuheng Lu, Chenfeng Xu, Xiaobao Wei, Xiaodong Xie, Masayoshi Tomizuka, Kurt Keutzer, Shanghang Zhang
In this paper, we address open-vocabulary 3D point-cloud detection by a dividing-and-conquering strategy, which involves: 1) developing a point-cloud detector that can learn a general representation for localizing various objects, and 2) connecting textual and point-cloud representations to enable the detector to classify novel object categories based on text prompting.
1 code implementation • CVPR 2023 • Yichen Xie, Han Lu, Junchi Yan, Xiaokang Yang, Masayoshi Tomizuka, Wei Zhan
We propose a novel method called ActiveFT for active finetuning task to select a subset of data distributing similarly with the entire unlabeled pool and maintaining enough diversity by optimizing a parametric model in the continuous space.
no code implementations • 24 Mar 2023 • Wei-Jer Chang, Chen Tang, Chenran Li, Yeping Hu, Masayoshi Tomizuka, Wei Zhan
To ensure that autonomous vehicles take safe and efficient maneuvers in different interactive traffic scenarios, we should be able to evaluate autonomous vehicles against reactive agents with different social characteristics in the simulation environment.
2 code implementations • 13 Feb 2023 • Haoyu Lu, Yuqi Huo, Guoxing Yang, Zhiwu Lu, Wei Zhan, Masayoshi Tomizuka, Mingyu Ding
Particularly, on the MSRVTT retrieval task, UniAdapter achieves 49. 7% recall@1 with 2. 2% model parameters, outperforming the latest competitors by 2. 0%.
1 code implementation • 3 Feb 2023 • Zhixuan Liang, Yao Mu, Mingyu Ding, Fei Ni, Masayoshi Tomizuka, Ping Luo
For example, AdaptDiffuser not only outperforms the previous art Diffuser by 20. 8% on Maze2D and 7. 5% on MuJoCo locomotion, but also adapts better to new tasks, e. g., KUKA pick-and-place, by 27. 9% without requiring additional expert data.
no code implementations • 2 Jan 2023 • Ran Tian, Masayoshi Tomizuka, Anca Dragan, Andrea Bajcsy
Interestingly, robot actions influence what this experience is, and therefore influence how people's internal models change.
1 code implementation • 21 Oct 2022 • Lingfeng Sun, Haichao Zhang, Wei Xu, Masayoshi Tomizuka
However, the gaps between contents and difficulties of different tasks bring us challenges on both which tasks should share the parameters and what parameters should be shared, as well as the optimization challenges due to parameter sharing.
1 code implementation • 5 Oct 2022 • Jinhyung Park, Chenfeng Xu, Shijia Yang, Kurt Keutzer, Kris Kitani, Masayoshi Tomizuka, Wei Zhan
While recent camera-only 3D detection methods leverage multiple timesteps, the limited history they use significantly hampers the extent to which temporal fusion can improve object perception.
Ranked #1 on Robust Camera Only 3D Object Detection on nuScenes-C
no code implementations • 1 Oct 2022 • Zheng Wu, Yichen Xie, Wenzhao Lian, Changhao Wang, Yanjiang Guo, Jianyu Chen, Stefan Schaal, Masayoshi Tomizuka
Experimental results demonstrate that our proposed method achieves policy generalization to unseen compositional tasks in a zero-shot manner.
no code implementations • 26 Sep 2022 • Philip Jacobson, Yiyang Zhou, Wei Zhan, Masayoshi Tomizuka, Ming C. Wu
In this work, we propose a novel approach Center Feature Fusion (CFF), in which we leverage center-based detection networks in both the camera and LiDAR streams to identify relevant object locations.
no code implementations • 9 Aug 2022 • Wei-Jer Chang, Yeping Hu, Chenran Li, Wei Zhan, Masayoshi Tomizuka
In this paper, we aim to provide a thorough stability analysis of the reactive simulation and propose a solution to enhance the stability.
no code implementations • 6 Aug 2022 • Juanwu Lu, Wei Zhan, Masayoshi Tomizuka, Yeping Hu
Results show significant performance degradation due to domain shift, and feature attribution provides insights to identify potential causes of these problems.
1 code implementation • 19 Jul 2022 • Guangming Wang, Yunzhe Hu, Zhe Liu, Yiyang Zhou, Masayoshi Tomizuka, Wei Zhan, Hesheng Wang
Our proposed model surpasses all existing methods by at least 38. 2% on FlyingThings3D dataset and 24. 7% on KITTI Scene Flow dataset for EPE3D metric.
no code implementations • 8 Jul 2022 • Akio Kodaira, Yiyang Zhou, Pengwei Zang, Wei Zhan, Masayoshi Tomizuka
With information from multiple input modalities, sensor fusion-based algorithms usually out-perform their single-modality counterparts in robotics.
no code implementations • 5 Jul 2022 • Yuheng Lu, Chenfeng Xu, Xiaobao Wei, Xiaodong Xie, Masayoshi Tomizuka, Kurt Keutzer, Shanghang Zhang
Current point-cloud detection methods have difficulty detecting the open-vocabulary objects in the real world, due to their limited generalization capability.
1 code implementation • 21 Apr 2022 • Chenfeng Xu, Tian Li, Chen Tang, Lingfeng Sun, Kurt Keutzer, Masayoshi Tomizuka, Alireza Fathi, Wei Zhan
It is hard to replicate these approaches in trajectory forecasting due to the lack of adequate trajectory data (e. g., 34K samples in the nuScenes dataset).
no code implementations • 19 Apr 2022 • Chen Tang, Wei Zhan, Masayoshi Tomizuka
Moreover, to properly evaluate an IBP model with offline datasets, we propose a Shapley-value-based metric to verify if the prediction model satisfies the inherent temporal independence of an interventional distribution.
no code implementations • 29 Mar 2022 • Xinghao Zhu, Siddarth Jain, Masayoshi Tomizuka, Jeroen van Baar
Vision-based tactile sensors typically utilize a deformable elastomer and a camera mounted above to provide high-resolution image observations of contacts.
no code implementations • 28 Mar 2022 • Lingfeng Sun, Chen Tang, Yaru Niu, Enna Sachdeva, Chiho Choi, Teruhisa Misu, Masayoshi Tomizuka, Wei Zhan
To address these issues, we propose a novel approach to avoid KL vanishing and induce an interpretable interactive latent space with pseudo labels.
1 code implementation • 17 Mar 2022 • Jinhyung Park, Chenfeng Xu, Yiyang Zhou, Masayoshi Tomizuka, Wei Zhan
While numerous 3D detection works leverage the complementary relationship between RGB images and point clouds, developments in the broader framework of semi-supervised object recognition remain uninfluenced by multi-modal fusion.
no code implementations • 5 Mar 2022 • Jiachen Li, Haiming Gang, Hengbo Ma, Masayoshi Tomizuka, Chiho Choi
We propose a novel approach for important object identification in egocentric driving scenarios with relational reasoning on the objects in the scene.
no code implementations • 10 Feb 2022 • Letian Wang, Yeping Hu, Liting Sun, Wei Zhan, Masayoshi Tomizuka, Changliu Liu
By mimicking humans' cognition model and semantic understanding during driving, we propose HATN, a hierarchical framework to generate high-quality, transferable, and adaptable predictions for driving behaviors in multi-agent dense-traffic environments.
no code implementations • 4 Jan 2022 • Hengbo Ma, Bike Zhang, Masayoshi Tomizuka, Koushil Sreenath
By embedding the optimization procedure of the exponential control barrier function based quadratic program (ECBF-QP) as a differentiable layer within a deep learning architecture, we propose a differentiable safety-critical control framework that enables generalization to new environments for high relative-degree systems with forward invariance guarantees.
no code implementations • CVPR 2022 • Hengbo Ma, Jiachen Li, Ramtin Hosseini, Masayoshi Tomizuka, Chiho Choi
Obtaining accurate and diverse human motion prediction is essential to many industrial applications, especially robotics and autonomous driving.
3 code implementations • 15 Dec 2021 • Yichen Xie, Masayoshi Tomizuka, Wei Zhan
Existing work follows a cumbersome pipeline that repeats the time-consuming model training and batch data selection multiple times.
no code implementations • 3 Dec 2021 • Yeping Hu, Xiaogang Jia, Masayoshi Tomizuka, Wei Zhan
In this paper, we aim to address the domain generalization problem for vehicle intention prediction tasks and a causal-based time series domain generalization (CTSDG) model is proposed.
no code implementations • NeurIPS 2021 • Chen Tang, Wei Zhan, Masayoshi Tomizuka
In this work, we argue that one of the typical formulations of VAEs in multi-agent modeling suffers from an issue we refer to as social posterior collapse, i. e., the model is prone to ignoring historical social context when predicting the future trajectory of an agent.
no code implementations • 9 Nov 2021 • Jinning Li, Chen Tang, Masayoshi Tomizuka, Wei Zhan
Reinforcement Learning (RL) has been shown effective in domains where the agent can learn policies by actively interacting with its operating environment.
no code implementations • 29 Sep 2021 • Rui Zhou, HongYu Zhou, Huidong Gao, Masayoshi Tomizuka, Jiachen Li, Zhuo Xu
Accurate, long-term forecasting of pedestrian trajectories in highly dynamic and interactive scenes is a long-standing challenge.
1 code implementation • 13 Sep 2021 • Zhao-Heng Yin, Lingfeng Sun, Hengbo Ma, Masayoshi Tomizuka, Wu-Jun Li
In this paper, we consider CDIL on a class of similar robots.
no code implementations • ICCV 2021 • Jiachen Li, Fan Yang, Hengbo Ma, Srikanth Malla, Masayoshi Tomizuka, Chiho Choi
Motion forecasting plays a significant role in various domains (e. g., autonomous driving, human-robot interaction), which aims to predict future motion sequences given a set of historical observations.
1 code implementation • 8 Jun 2021 • Chenfeng Xu, Shijia Yang, Tomer Galanti, Bichen Wu, Xiangyu Yue, Bohan Zhai, Wei Zhan, Peter Vajda, Kurt Keutzer, Masayoshi Tomizuka
We discover that we can indeed use the same architecture and pretrained weights of a neural net model to understand both images and point-clouds.
no code implementations • 5 Jun 2021 • Defu Cao, Jiachen Li, Hengbo Ma, Masayoshi Tomizuka
To this end, we propose a Spectral Temporal Graph Neural Network (SpecTGNN), which can capture inter-agent correlations and temporal dependency simultaneously in frequency domain in addition to time domain.
no code implementations • 4 Jun 2021 • Zhuo Xu, Masayoshi Tomizuka
In this extended abstract, we investigate the design of learning representation for human intention inference.
no code implementations • 22 Mar 2021 • Jun Ma, Zilong Cheng, Haiyue Zhu, Xiaocong Li, Masayoshi Tomizuka, Tong Heng Lee
Magnetic levitation positioning technology has attracted considerable research efforts and dedicated attention due to its extremely attractive features.
no code implementations • 7 Mar 2021 • Ran Tian, Masayoshi Tomizuka, Liting Sun
In this work, we advocate that humans are bounded rational and have different intelligence levels when reasoning about others' decision-making process, and such an inherent and latent characteristic should be accounted for in reward learning algorithms.
1 code implementation • 6 Mar 2021 • Di Feng, Yiyang Zhou, Chenfeng Xu, Masayoshi Tomizuka, Wei Zhan
Detecting dynamic objects and predicting static road information such as drivable areas and ground heights are crucial for safe autonomous driving.
no code implementations • 1 Mar 2021 • Zhao-Heng Yin, Lingfeng Sun, Liting Sun, Masayoshi Tomizuka, Wei Zhan
Experiments show that our model can generate diverse interactions in various scenarios.
no code implementations • 23 Feb 2021 • Chen Tang, Nishan Srishankar, Sujitha Martin, Masayoshi Tomizuka
Explainability is essential for autonomous vehicles and other robotics systems interacting with humans and other objects during operation.
no code implementations • 18 Feb 2021 • Jiachen Li, Hengbo Ma, Zhihao Zhang, Jinning Li, Masayoshi Tomizuka
Due to the existence of frequent interactions and uncertainty in the scene evolution, it is desired for the prediction system to enable relational reasoning on different entities and provide a distribution of future trajectories for each agent.
no code implementations • 29 Jan 2021 • Shiyu Jin, Xinghao Zhu, Changhao Wang, Masayoshi Tomizuka
Peg-in-hole assembly is a challenging contact-rich manipulation task.
Robotics
no code implementations • 17 Jan 2021 • Jinning Li, Liting Sun, Jianyu Chen, Masayoshi Tomizuka, Wei Zhan
To address this challenge, we propose a hierarchical behavior planning framework with a set of low-level safe controllers and a high-level reinforcement learning algorithm (H-CtRL) as a coordinator for the low-level controllers.
no code implementations • ICCV 2021 • Bichen Wu, Chenfeng Xu, Xiaoliang Dai, Alvin Wan, Peizhao Zhang, Zhicheng Yan, Masayoshi Tomizuka, Joseph E. Gonzalez, Kurt Keutzer, Peter Vajda
A recent trend in computer vision is to replace convolutions with transformers.
no code implementations • 18 Dec 2020 • Di Feng, Zining Wang, Yiyang Zhou, Lars Rosenbaum, Fabian Timm, Klaus Dietmayer, Masayoshi Tomizuka, Wei Zhan
As a result, an in-depth evaluation among different object detection methods remains challenging, and the training process of object detectors is sub-optimal, especially in probabilistic object detection.
6 code implementations • CVPR 2021 • Peize Sun, Rufeng Zhang, Yi Jiang, Tao Kong, Chenfeng Xu, Wei Zhan, Masayoshi Tomizuka, Lei LI, Zehuan Yuan, Changhu Wang, Ping Luo
In our method, however, a fixed sparse set of learned object proposals, total length of $N$, are provided to object recognition head to perform classification and location.
Ranked #5 on 2D Object Detection on CeyMo
no code implementations • 23 Nov 2020 • Zhuo Xu, Wenhao Yu, Alexander Herzog, Wenlong Lu, Chuyuan Fu, Masayoshi Tomizuka, Yunfei Bai, C. Karen Liu, Daniel Ho
General contact-rich manipulation problems are long-standing challenges in robotics due to the difficulty of understanding complicated contact physics.
no code implementations • 4 Nov 2020 • Xiaosong Jia, Liting Sun, Masayoshi Tomizuka, Wei Zhan
We find three interpretable patterns of interactions, bringing insights for driver behavior representation, modeling and comprehension.
no code implementations • 28 Oct 2020 • Letian Wang, Liting Sun, Masayoshi Tomizuka, Wei Zhan
It allows the AVs to infer the characteristics of other road users online and generate behaviors optimizing not only their own rewards, but also their courtesy to others, and their confidence regarding the prediction uncertainties.
no code implementations • 3 Sep 2020 • Ran Tian, Liting Sun, Masayoshi Tomizuka
Classical game-theoretic approaches for multi-agent systems in both the forward policy design problem and the inverse reward learning problem often make strong rationality assumptions: agents perfectly maximize expected utilities under uncertainties.
no code implementations • 20 Aug 2020 • Liting Sun, Zheng Wu, Hengbo Ma, Masayoshi Tomizuka
In human-robot interaction (HRI) systems, such as autonomous vehicles, understanding and representing human behavior are important.
no code implementations • 22 Jun 2020 • Zheng Wu, Liting Sun, Wei Zhan, Chenyu Yang, Masayoshi Tomizuka
Different from existing IRL algorithms, by introducing an efficient continuous-domain trajectory sampler, the proposed algorithm can directly learn the reward functions in the continuous domain while considering the uncertainties in demonstrated trajectories from human drivers.
no code implementations • 22 Jun 2020 • Hujie Pan, Zining Wang, Wei Zhan, Masayoshi Tomizuka
In this paper, we propose a novel form of the loss function to increase the performance of LiDAR-based 3d object detection and obtain more explainable and convincing uncertainty for the prediction.
no code implementations • 9 Jun 2020 • Saman Fahandezh-Saadi, Masayoshi Tomizuka
Rectifier (ReLU) deep neural networks (DNN) and their connection with piecewise affine (PWA) functions is analyzed.
8 code implementations • 5 Jun 2020 • Bichen Wu, Chenfeng Xu, Xiaoliang Dai, Alvin Wan, Peizhao Zhang, Zhicheng Yan, Masayoshi Tomizuka, Joseph Gonzalez, Kurt Keutzer, Peter Vajda
In this work, we challenge this paradigm by (a) representing images as semantic visual tokens and (b) running transformers to densely model token relationships.
no code implementations • 7 Apr 2020 • Yeping Hu, Wei Zhan, Masayoshi Tomizuka
Accurately predicting the possible behaviors of traffic participants is an essential capability for autonomous vehicles.
3 code implementations • ECCV 2020 • Chenfeng Xu, Bichen Wu, Zining Wang, Wei Zhan, Peter Vajda, Kurt Keutzer, Masayoshi Tomizuka
Using standard convolutions to process such LiDAR images is problematic, as convolution filters pick up local features that are only active in specific regions in the image.
Ranked #24 on 3D Semantic Segmentation on SemanticKITTI
no code implementations • NeurIPS 2020 • Jiachen Li, Fan Yang, Masayoshi Tomizuka, Chiho Choi
In this paper, we propose a generic trajectory forecasting framework (named EvolveGraph) with explicit relational structure recognition and prediction via latent interaction graphs among multiple heterogeneous, interactive agents.
Ranked #12 on Trajectory Prediction on Stanford Drone
2 code implementations • 21 Mar 2020 • Jianyu Chen, Zhuo Xu, Masayoshi Tomizuka
Current autonomous driving systems are composed of a perception system and a decision system.
no code implementations • 7 Mar 2020 • Zining Wang, Di Feng, Yiyang Zhou, Lars Rosenbaum, Fabian Timm, Klaus Dietmayer, Masayoshi Tomizuka, Wei Zhan
Based on the spatial distribution, we further propose an extension of IoU, called the Jaccard IoU (JIoU), as a new evaluation metric that incorporates label uncertainty.
no code implementations • 14 Feb 2020 • Jiachen Li, Hengbo Ma, Zhihao Zhang, Masayoshi Tomizuka
Effective understanding of the environment and accurate trajectory prediction of surrounding dynamic obstacles are indispensable for intelligent mobile systems (like autonomous vehicles and social robots) to achieve safe and high-quality planning when they navigate in highly interactive and crowded scenarios.
4 code implementations • 23 Jan 2020 • Jianyu Chen, Shengbo Eben Li, Masayoshi Tomizuka
A sequential latent environment model is introduced and learned jointly with the reinforcement learning process.
2 code implementations • 20 Dec 2019 • Chenfeng Xu, Dingkang Liang, Yongchao Xu, Song Bai, Wei Zhan, Xiang Bai, Masayoshi Tomizuka
A major issue is that the density map on dense regions usually accumulates density values from a number of nearby Gaussian blobs, yielding different large density values on a small set of pixels.
no code implementations • 19 Dec 2019 • Weisong Wen, Yiyang Zhou, Guohao Zhang, Saman Fahandezh-Saadi, Xiwei Bai, Wei Zhan, Masayoshi Tomizuka, Li-Ta Hsu
Mapping and localization is a critical module of autonomous driving, and significant achievements have been reached in this field.
no code implementations • 25 Nov 2019 • Saman Fahandezh-Saadi, Di Wang, Masayoshi Tomizuka
This paper presents a robust probabilistic point registration method for estimating the rigid transformation (i. e. rotation matrix and translation vector) between two pointcloud dataset.
no code implementations • 11 Nov 2019 • Chen Tang, Jianyu Chen, Masayoshi Tomizuka
Current methods for long-term trajectory prediction cannot guarantee the physical feasibility of predicted distribution.
no code implementations • 17 Oct 2019 • Huidong Gao, Yi Ouyang, Masayoshi Tomizuka
In this paper, we propose a combined prediction model and an online learning framework for planar push prediction.
no code implementations • 30 Sep 2019 • Wei Zhan, Liting Sun, Di Wang, Haojie Shi, Aubrey Clausse, Maximilian Naumann, Julius Kummerle, Hendrik Konigshof, Christoph Stiller, Arnaud de La Fortelle, Masayoshi Tomizuka
3) The driving behavior is highly interactive and complex with adversarial and cooperative motions of various traffic participants.
no code implementations • 27 Sep 2019 • Kiwoo Shin, Masayoshi Tomizuka
Our approach can improve a 3D object detection performance by predicting more precise 3D bounding box coordinates.
no code implementations • 23 Aug 2019 • Jiachen Li, Wei Zhan, Yeping Hu, Masayoshi Tomizuka
The framework can incorporate an arbitrary prediction model as the implicit proposal distribution of the CMSMC method.
no code implementations • 23 Jul 2019 • Yeping Hu, Liting Sun, Masayoshi Tomizuka
Both rational and irrational behaviors exist, and the autonomous vehicles need to be aware of this in their prediction module.
no code implementations • 19 Jul 2019 • Liting Sun, Wei Zhan, Yeping Hu, Masayoshi Tomizuka
Hence, the goal of this work is to formulate the human drivers' behavior generation model with CPT so that some ``irrational'' behavior or decisions of human can be better captured and predicted.
no code implementations • 5 May 2019 • Jiachen Li, Hengbo Ma, Masayoshi Tomizuka
Effective understanding of the environment and accurate trajectory prediction of surrounding dynamic obstacles are critical for intelligent systems such as autonomous vehicles and wheeled mobile robotics navigating in complex scenarios to achieve safe and high-quality decision making, motion planning and control.
Ranked #14 on Trajectory Prediction on Stanford Drone
no code implementations • 2 May 2019 • Jiachen Li, Hengbo Ma, Wei Zhan, Masayoshi Tomizuka
In order to tackle the task of probabilistic prediction for multiple, interactive entities, we propose a coordination and trajectory prediction system (CTPS), which has a hierarchical structure including a macro-level coordination recognition module and a micro-level subtle pattern prediction module which solves a probabilistic generation task.
no code implementations • 2 May 2019 • Liting Sun, Wei Zhan, Ching-Yao Chan, Masayoshi Tomizuka
The uncertainties can come either from sensor limitations such as occlusions and limited sensor range, or from probabilistic prediction of other road participants, or from unknown social behavior in a new area.
2 code implementations • 20 Apr 2019 • Jianyu Chen, Bodi Yuan, Masayoshi Tomizuka
Urban autonomous driving decision making is challenging due to complex road geometry and multi-agent interactions.
no code implementations • 15 Apr 2019 • Yongxiang Fan, Xinghao Zhu, Masayoshi Tomizuka
Searching precision grasps on the object represented by point cloud, is challenging due to the complex object shape, high-dimensionality, collision and undesired properties of the sensing and positioning.
Robotics
no code implementations • 12 Apr 2019 • Yeping Hu, Alireza Nakhaei, Masayoshi Tomizuka, Kikuo Fujimura
In this paper, we proposed an interaction-aware decision making with adaptive strategies (IDAS) approach that can let the autonomous vehicle negotiate the road with other drivers by leveraging their cooperativeness under merging scenarios.
no code implementations • 4 Apr 2019 • Jiachen Li, Hengbo Ma, Masayoshi Tomizuka
In order to enable high-quality decision making and motion planning of intelligent systems such as robotics and autonomous vehicles, accurate probabilistic predictions for surrounding interactive objects is a crucial prerequisite.
no code implementations • 22 Mar 2019 • Yeping Hu, Wei Zhan, Liting Sun, Masayoshi Tomizuka
The proposed method is based on a generative model and is capable of jointly predicting sequential motions of each pair of interacting agents.
no code implementations • 28 Feb 2019 • Yongxiang Fan, Masayoshi Tomizuka
The framework includes a multi-dimensional iterative surface fitting (MDISF) for grasp planning and a grasp trajectory optimization (GTO) for grasp imagination.
Robotics
no code implementations • 7 Dec 2018 • Zhuo Xu, Chen Tang, Masayoshi Tomizuka
Although deep reinforcement learning (deep RL) methods have lots of strengths that are favorable if applied to autonomous driving, real deep RL applications in autonomous driving have been slowed down by the modeling gap between the source (training) domain and the target (deployment) domain.
no code implementations • 9 Nov 2018 • Kiwoo Shin, Youngwook Paul Kwon, Masayoshi Tomizuka
We present RoarNet, a new approach for 3D object detection from a 2D image and 3D Lidar point clouds.
Ranked #3 on Object Detection on KITTI Cars Easy
no code implementations • 30 Oct 2018 • Yeping Hu, Wei Zhan, Masayoshi Tomizuka
In a given scenario, simultaneously and accurately predicting every possible interaction of traffic participants is an important capability for autonomous vehicles.
no code implementations • 23 Sep 2018 • Yongxiang Fan, Hsien-Chung Lin, Te Tang, Masayoshi Tomizuka
In this paper, we propose a learning framework to plan robust grasps for customized grippers in real-time.
no code implementations • 23 Sep 2018 • Yongxiang Fan, Jieliang Luo, Masayoshi Tomizuka
The framework combines both the supervised learning and the reinforcement learning.
no code implementations • 10 Sep 2018 • Wei Zhan, Liting Sun, Yeping Hu, Jiachen Li, Masayoshi Tomizuka
Modified methods based on PGM, NN and IRL are provided to generate probabilistic reaction predictions in an exemplar scenario of nudging from a highway ramp.
no code implementations • 9 Sep 2018 • Liting Sun, Wei Zhan, Masayoshi Tomizuka
To safely and efficiently interact with other road participants, AVs have to accurately predict the behavior of surrounding vehicles and plan accordingly.
no code implementations • 9 Sep 2018 • Jiachen Li, Hengbo Ma, Wei Zhan, Masayoshi Tomizuka
Accurate and robust recognition and prediction of traffic situation plays an important role in autonomous driving, which is a prerequisite for risk assessment and effective decision making.
1 code implementation • 12 Aug 2018 • Changliu Liu, Masayoshi Tomizuka
Human-robot interactions have been recognized to be a key element of future industrial collaborative robots (co-robots).
Robotics Systems and Control
no code implementations • 8 Aug 2018 • Liting Sun, Wei Zhan, Masayoshi Tomizuka, Anca D. Dragan
Such a courtesy term enables the robot car to be aware of possible irrationality of the human behavior, and plan accordingly.
no code implementations • 10 Apr 2018 • Yeping Hu, Wei Zhan, Masayoshi Tomizuka
Accurately predicting the possible behaviors of traffic participants is an essential capability for future autonomous vehicles.
no code implementations • 30 Mar 2018 • Yongxiang Fan, Hsien-Chung Lin, Te Tang, Masayoshi Tomizuka
The proposed algorithm is able to consider the structural constraints of the gripper and plan optimal grasps in real-time.
Robotics
no code implementations • 24 Nov 2017 • Zhuo Xu, Haonan Chang, Masayoshi Tomizuka
We propose the cascade attribute learning network (CALNet), which can learn attributes in a control task separately and assemble them together.
no code implementations • 17 Nov 2017 • Zining Wang, Wei Zhan, Masayoshi Tomizuka
The fusion method shows particular benefit for detection of pedestrians in the bird view compared to other fusion-based object detection networks.
1 code implementation • 2 Sep 2017 • Changliu Liu, Chung-Yen Lin, Masayoshi Tomizuka
The idea is to find a convex feasible set for the original problem and iteratively solve a sequence of subproblems using the convex constraints.
Optimization and Control Robotics
no code implementations • 9 Jul 2017 • Liting Sun, Cheng Peng, Wei Zhan, Masayoshi Tomizuka
For safe and efficient planning and control in autonomous driving, we need a driving policy which can achieve desirable driving quality in long-term horizon with guaranteed safety and feasibility.