arXiv subjects

Masayoshi Tomizuka

Publications and source records attributed to Masayoshi Tomizuka.

At least 253 records · Page 14Linked to original sources

arXiv2021

SqueezeSegV3: Spatially-Adaptive Convolution for Efficient Point-Cloud Segmentation

LiDAR point-cloud segmentation is an important problem for many applications. For large-scale point cloud segmentation, the \textit{de facto} method is to project a 3D point cloud to get a 2D LiDAR image and use convolutions to process it. Despite the similarity between regular RGB and LiDAR images, we discover that the feature distribution of LiDAR images changes drastically at different image locations. 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. As a result, the capacity of the network is under-utilized and the segmentation performance decreases. To fix this, we propose Spatially-Adaptive Convolution (SAC) to adopt different filters for different locations according to the input image. SAC can be computed efficiently since it can be implemented as a series of element-wise multiplications, im2col, and standard convolution. It is a general framework such that several previous methods can be seen as special cases of SAC. Using SAC, we build SqueezeSegV3 for LiDAR point-cloud segmentation and outperform all previous published methods by at least 3.7% mIoU on the SemanticKITTI benchmark with comparable inference speed.

cs.CV↗

arXiv2021

You Only Group Once: Efficient Point-Cloud Processing with Token Representation and Relation Inference Module

3D point-cloud-based perception is a challenging but crucial computer vision task. A point-cloud consists of a sparse, unstructured, and unordered set of points. To understand a point-cloud, previous point-based methods, such as PointNet++, extract visual features through hierarchically aggregation of local features. However, such methods have several critical limitations: 1) Such methods require several sampling and grouping operations, which slow down the inference speed. 2) Such methods spend an equal amount of computation on each points in a point-cloud, though many of points are redundant. 3) Such methods aggregate local features together through downsampling, which leads to information loss and hurts the perception performance. To overcome these challenges, we propose a novel, simple, and elegant deep learning model called YOGO (You Only Group Once). Compared with previous methods, YOGO only needs to sample and group a point-cloud once, so it is very efficient. Instead of operating on points, YOGO operates on a small number of tokens, each of which summarizes the point features in a sub-region. This allows us to avoid computing on the redundant points and thus boosts efficiency.Moreover, YOGO preserves point-wise features by projecting token features to point features although the computation is performed on tokens. This avoids information loss and can improve point-wise perception performance. We conduct thorough experiments to demonstrate that YOGO achieves at least 3.0x speedup over point-based baselines while delivering competitive classification and segmentation performance on the ModelNet, ShapeNetParts and S3DIS datasets.

cs.RO↗

arXiv2021

Bounded Risk-Sensitive Markov Games: Forward Policy Design and Inverse Reward Learning with Iterative Reasoning and Cumulative Prospect Theory

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. Such assumptions, however, substantially mismatch with observed humans' behaviors such as satisficing with sub-optimal, risk-seeking, and loss-aversion decisions. In this paper, we investigate the problem of bounded risk-sensitive Markov Game (BRSMG) and its inverse reward learning problem for modeling human realistic behaviors and learning human behavioral models. Drawing on iterative reasoning models and cumulative prospect theory, we embrace that humans have bounded intelligence and maximize risk-sensitive utilities in BRSMGs. Convergence analysis for both the forward policy design and the inverse reward learning problems are established under the BRSMG framework. We validate the proposed forward policy design and inverse reward learning algorithms in a navigation scenario. The results show that the behaviors of agents demonstrate both risk-averse and risk-seeking characteristics. Moreover, in the inverse reward learning task, the proposed bounded risk-sensitive inverse learning algorithm outperforms a baseline risk-neutral inverse learning algorithm by effectively recovering not only more accurate reward values but also the intelligence levels and the risk-measure parameters given demonstrations of agents' interactive behaviors.

cs.LG↗

arXiv2021

Learning Human Rewards by Inferring Their Latent Intelligence Levels in Multi-Agent Games: A Theory-of-Mind Approach with Application to Driving Data

Reward function, as an incentive representation that recognizes humans' agency and rationalizes humans' actions, is particularly appealing for modeling human behavior in human-robot interaction. Inverse Reinforcement Learning is an effective way to retrieve reward functions from demonstrations. However, it has always been challenging when applying it to multi-agent settings since the mutual influence between agents has to be appropriately modeled. To tackle this challenge, previous work either exploits equilibrium solution concepts by assuming humans as perfectly rational optimizers with unbounded intelligence or pre-assigns humans' interaction strategies a priori. 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. Hence, we exploit such insights from Theory-of-Mind and propose a new multi-agent Inverse Reinforcement Learning framework that reasons about humans' latent intelligence levels during learning. We validate our approach in both zero-sum and general-sum games with synthetic agents and illustrate a practical application to learning human drivers' reward functions from real driving data. We compare our approach with two baseline algorithms. The results show that by reasoning about humans' latent intelligence levels, the proposed approach has more flexibility and capability to retrieve reward functions that explain humans' driving behaviors better.

cs.AI↗

arXiv2021

A Simple and Efficient Multi-task Network for 3D Object Detection and Road Understanding

Detecting dynamic objects and predicting static road information such as drivable areas and ground heights are crucial for safe autonomous driving. Previous works studied each perception task separately, and lacked a collective quantitative analysis. In this work, we show that it is possible to perform all perception tasks via a simple and efficient multi-task network. Our proposed network, LidarMTL, takes raw LiDAR point cloud as inputs, and predicts six perception outputs for 3D object detection and road understanding. The network is based on an encoder-decoder architecture with 3D sparse convolution and deconvolution operations. Extensive experiments verify the proposed method with competitive accuracies compared to state-of-the-art object detectors and other task-specific networks. LidarMTL is also leveraged for online localization. Code and pre-trained model have been made available at https://github.com/frankfengdi/LidarMTL.

cs.CV↗

arXiv2021

Diverse Critical Interaction Generation for Planning and Planner Evaluation

Generating diverse and comprehensive interacting agents to evaluate the decision-making modules is essential for the safe and robust planning of autonomous vehicles~(AV). Due to efficiency and safety concerns, most researchers choose to train interactive adversary~(competitive or weakly competitive) agents in simulators and generate test cases to interact with evaluated AVs. However, most existing methods fail to provide both natural and critical interaction behaviors in various traffic scenarios. To tackle this problem, we propose a styled generative model RouteGAN that generates diverse interactions by controlling the vehicles separately with desired styles. By altering its style coefficients, the model can generate trajectories with different safety levels serve as an online planner. Experiments show that our model can generate diverse interactions in various scenarios. We evaluate different planners with our model by testing their collision rate in interaction with RouteGAN planners of multiple critical levels.

cs.RO↗

arXiv2021

Spatio-Temporal Graph Dual-Attention Network for Multi-Agent Prediction and Tracking

An effective understanding of the environment and accurate trajectory prediction of surrounding dynamic obstacles are indispensable for intelligent mobile systems (e.g. autonomous vehicles and social robots) to achieve safe and high-quality planning when they navigate in highly interactive and crowded scenarios. 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. In this paper, we propose a generic generative neural system (called STG-DAT) for multi-agent trajectory prediction involving heterogeneous agents. The system takes a step forward to explicit interaction modeling by incorporating relational inductive biases with a dynamic graph representation and leverages both trajectory and scene context information. We also employ an efficient kinematic constraint layer applied to vehicle trajectory prediction. The constraint not only ensures physical feasibility but also enhances model performance. Moreover, the proposed prediction model can be easily adopted by multi-target tracking frameworks. The tracking accuracy proves to be improved by empirical results. The proposed system is evaluated on three public benchmark datasets for trajectory prediction, where the agents cover pedestrians, cyclists and on-road vehicles. The experimental results demonstrate that our model achieves better performance than various baseline approaches in terms of prediction and tracking accuracy.

cs.CV↗

arXiv2021

Learning Variable Impedance Control via Inverse Reinforcement Learning for Force-Related Tasks

Many manipulation tasks require robots to interact with unknown environments. In such applications, the ability to adapt the impedance according to different task phases and environment constraints is crucial for safety and performance. Although many approaches based on deep reinforcement learning (RL) and learning from demonstration (LfD) have been proposed to obtain variable impedance skills on contact-rich manipulation tasks, these skills are typically task-specific and could be sensitive to changes in task settings. This paper proposes an inverse reinforcement learning (IRL) based approach to recover both the variable impedance policy and reward function from expert demonstrations. We explore different action space of the reward functions to achieve a more general representation of expert variable impedance skills. Experiments on two variable impedance tasks (Peg-in-Hole and Cup-on-Plate) were conducted in both simulations and on a real FANUC LR Mate 200iD/7L industrial robot. The comparison results with behavior cloning and force-based IRL proved that the learned reward function in the gain action space has better transferability than in the force space. Experiment videos are available at https://msc.berkeley.edu/research/impedance-irl.html.

cs.RO↗

arXiv2021

Feedback-based Digital Higher-order Terminal Sliding Mode for 6-DOF Industrial Manipulators

The precise motion control of a multi-degree of freedom~(DOF) robot manipulator is always challenging due to its nonlinear dynamics, disturbances, and uncertainties. Because most manipulators are controlled by digital signals, a novel higher-order sliding mode controller in the discrete-time form with time delay estimation is proposed in this paper. The dynamic model of the manipulator used in the design allows proper handling of nonlinearities, uncertainties and disturbances involved in the problem. Specifically, parametric uncertainties and disturbances are handled by the time delay estimation and the nonlinearity of the manipulator is addressed by the feedback structure of the controller. The combination of terminal sliding mode surface and higher-order control scheme in the controller guarantees a fast response with a small chattering amplitude. Moreover, the controller is designed with a modified sliding mode surface and variable-gain structure, so that the performance of the controller is further enhanced. We also analyze the condition to guarantee the stability of the closed-loop system in this paper. Finally, the simulation and experimental results prove that the proposed control scheme has a precise performance in a robot manipulator system.

cs.RO↗

arXiv2021

Practical Fractional-Order Variable-Gain Super-Twisting Control with Application to Wafer Stages of Photolithography Systems

In this paper, a practical fractional-order variable-gain super-twisting algorithm (PFVSTA) is proposed to improve the tracking performance of wafer stages for semiconductor manufacturing. Based on the sliding mode control (SMC), the proposed PFVSTA enhances the tracking performance from three aspects: 1) alleviating the chattering phenomenon via super-twisting algorithm and a novel fractional-order sliding surface~(FSS) design, 2) improving the dynamics of states on the sliding surface with fast response and small overshoots via the designed novel FSS and 3) compensating for disturbances via variable-gain control law. Based on practical conditions, this paper analyzes the stability of the controller and illustrates the theoretical principle to compensate for the uncertainties caused by accelerations. Moreover, numerical simulations prove the effectiveness of the proposed sliding surface and control scheme, and they are in agreement with the theoretical analysis. Finally, practice-based comparative experiments are conducted. The results show that the proposed PFVSTA can achieve much better tracking performance than the conventional methods from various perspectives.

eess.SY↗

arXiv2021

Precise Motion Control of Wafer Stages via Adaptive Neural Network and Fractional-Order Super-Twisting Algorithm

To obtain precise motion control of wafer stages, an adaptive neural network and fractional-order super-twisting control strategy is proposed. Based on sliding mode control (SMC), the proposed controller aims to address two challenges in SMC: 1) reducing the chattering phenomenon, and 2) attenuating the influence of model uncertainties and disturbances. For the first challenge, a fractional-order terminal sliding mode surface and a super-twisting algorithm are integrated into the SMC design. To attenuate uncertainties and disturbances, an add-on control structure based on the radial basis function (RBF) neural network is introduced. Stability analysis of the closed-loop control system is provided. Finally, experiments on a wafer stage testbed system are conducted, which proves that the proposed controller can robustly improve the tracking performance in the presence of uncertainties and disturbances compared to conventional and previous controllers.

eess.SY↗

arXiv2021

Contact Pose Identification for Peg-in-Hole Assembly under Uncertainties

Peg-in-hole assembly is a challenging contact-rich manipulation task. There is no general solution to identify the relative position and orientation between the peg and the hole. In this paper, we propose a novel method to classify the contact poses based on a sequence of contact measurements. When the peg contacts the hole with pose uncertainties, a tilt-then-rotate strategy is applied, and the contacts are measured as a group of patterns to encode the contact pose. A convolutional neural network (CNN) is trained to classify the contact poses according to the patterns. In the end, an admittance controller guides the peg towards the error direction and finishes the peg-in-hole assembly. Simulations and experiments are provided to show that the proposed method can be applied to the peg-in-hole assembly of different geometries. We also demonstrate the ability to alleviate the sim-to-real gap.

cs.RO↗

arXiv2021

Interaction-Aware Behavior Planning for Autonomous Vehicles Validated with Real Traffic Data

Autonomous vehicles (AVs) need to interact with other traffic participants who can be either cooperative or aggressive, attentive or inattentive. Such different characteristics can lead to quite different interactive behaviors. Hence, to achieve safe and efficient autonomous driving, AVs need to be aware of such uncertainties when they plan their own behaviors. In this paper, we formulate such a behavior planning problem as a partially observable Markov Decision Process (POMDP) where the cooperativeness of other traffic participants is treated as an unobservable state. Under different cooperativeness levels, we learn the human behavior models from real traffic data via the principle of maximum likelihood. Based on that, the POMDP problem is solved by Monte-Carlo Tree Search. We verify the proposed algorithm in both simulations and real traffic data on a lane change scenario, and the results show that the proposed algorithm can successfully finish the lane changes without collisions.

cs.RO↗

arXiv2021

Optimal Decentralized Control for Uncertain Systems by Symmetric Gauss-Seidel Semi-Proximal ALM

The H2 guaranteed cost decentralized control problem is investigated in this work. More specifically, on the basis of an appropriate H2 re-formulation that we put in place, the optimal control problem in the presence of parameter uncertainties is then suitably characterized by convex restriction and solved in parameter space. It is shown that a set of stabilizing decentralized controller gains for the uncertain system is parameterized in a convex set through appropriate convex restriction, and then an approximated conic optimization problem is constructed. This facilitates the use of the symmetric Gauss-Seidel (sGS) semi-proximal augmented Lagrangian method (ALM), which attains high computational effectiveness. A comprehensive analysis is given on the application of the approach in solving the optimal decentralized control problem; and subsequently, the preserved decentralized structure, robust stability, and robust performance are all suitably guaranteed with the proposed methodology. Furthermore, an illustrative example is presented to demonstrate the effectiveness of the proposed optimization approach.

math.OC↗

arXiv2020

epBRM: Improving a Quality of 3D Object Detection using End Point Box Regression Module

We present an endpoint box regression module(epBRM), which is designed for predicting precise 3D bounding boxes using raw LiDAR 3D point clouds. The proposed epBRM is built with sequence of small networks and is computationally lightweight. Our approach can improve a 3D object detection performance by predicting more precise 3D bounding box coordinates. The proposed approach requires 40 minutes of training to improve the detection performance. Moreover, epBRM imposes less than 12ms to network inference time for up-to 20 objects. The proposed approach utilizes a spatial transformation mechanism to simplify the box regression task. Adopting spatial transformation mechanism into epBRM makes it possible to improve the quality of detection with a small sized network. We conduct in-depth analysis of the effect of various spatial transformation mechanisms applied on raw LiDAR 3D point clouds. We also evaluate the proposed epBRM by applying it to several state-of-the-art 3D object detection systems. We evaluate our approach on KITTI dataset, a standard 3D object detection benchmark for autonomous vehicles. The proposed epBRM enhances the overlaps between ground truth bounding boxes and detected bounding boxes, and improves 3D object detection. Our proposed method evaluated in KITTI test server outperforms current state-of-the-art approaches.

cs.CV↗

arXiv2020

Labels Are Not Perfect: Inferring Spatial Uncertainty in Object Detection

The availability of many real-world driving datasets is a key reason behind the recent progress of object detection algorithms in autonomous driving. However, there exist ambiguity or even failures in object labels due to error-prone annotation process or sensor observation noise. Current public object detection datasets only provide deterministic object labels without considering their inherent uncertainty, as does the common training process or evaluation metrics for object detectors. 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. In this work, we infer the uncertainty in bounding box labels from LiDAR point clouds based on a generative model, and define a new representation of the probabilistic bounding box through a spatial uncertainty distribution. Comprehensive experiments show that the proposed model reflects complex environmental noises in LiDAR perception and the label quality. Furthermore, we propose Jaccard IoU (JIoU) as a new evaluation metric that extends IoU by incorporating label uncertainty. We conduct an in-depth comparison among several LiDAR-based object detectors using the JIoU metric. Finally, we incorporate the proposed label uncertainty in a loss function to train a probabilistic object detector and to improve its detection accuracy. We verify our proposed methods on two public datasets (KITTI, Waymo), as well as on simulation data. Code is released at https://bit.ly/2W534yo.

cs.CV↗

arXiv2020

COCOI: Contact-aware Online Context Inference for Generalizable Non-planar Pushing

General contact-rich manipulation problems are long-standing challenges in robotics due to the difficulty of understanding complicated contact physics. Deep reinforcement learning (RL) has shown great potential in solving robot manipulation tasks. However, existing RL policies have limited adaptability to environments with diverse dynamics properties, which is pivotal in solving many contact-rich manipulation tasks. In this work, we propose Contact-aware Online COntext Inference (COCOI), a deep RL method that encodes a context embedding of dynamics properties online using contact-rich interactions. We study this method based on a novel and challenging non-planar pushing task, where the robot uses a monocular camera image and wrist force torque sensor reading to push an object to a goal location while keeping it upright. We run extensive experiments to demonstrate the capability of COCOI in a wide range of settings and dynamics properties in simulation, and also in a sim-to-real transfer scenario on a real robot (Video: https://youtu.be/nrmJYksh1Kc)

cs.RO↗

arXiv2020

Visual Transformers: Token-based Image Representation and Processing for Computer Vision

Computer vision has achieved remarkable success by (a) representing images as uniformly-arranged pixel arrays and (b) convolving highly-localized features. However, convolutions treat all image pixels equally regardless of importance; explicitly model all concepts across all images, regardless of content; and struggle to relate spatially-distant concepts. In this work, we challenge this paradigm by (a) representing images as semantic visual tokens and (b) running transformers to densely model token relationships. Critically, our Visual Transformer operates in a semantic token space, judiciously attending to different image parts based on context. This is in sharp contrast to pixel-space transformers that require orders-of-magnitude more compute. Using an advanced training recipe, our VTs significantly outperform their convolutional counterparts, raising ResNet accuracy on ImageNet top-1 by 4.6 to 7 points while using fewer FLOPs and parameters. For semantic segmentation on LIP and COCO-stuff, VT-based feature pyramid networks (FPN) achieve 0.35 points higher mIoU while reducing the FPN module's FLOPs by 6.5x.

cs.CV↗