Publishing type：Research paper (scientific journal)  

Rich semantic information extraction plays a vital role on next-generation
intelligent vehicles. Currently there is great amount of research focusing on
fundamental applications such as 6D pose detection, road scene semantic
segmentation, etc. And this provides us a great opportunity to think about how
shall these data be organized and exploited.
In this paper we propose road scene graph,a special scene-graph for
intelligent vehicles. Different to classical data representation, this graph
provides not only object proposals but also their pair-wise relationships. By
organizing them in a topological graph, these data are explainable,
fully-connected, and could be easily processed by GCNs (Graph Convolutional
Networks). Here we apply scene graph on roads using our Road Scene Graph
dataset, including the basic graph prediction model. This work also includes
experimental evaluations using the proposed model.

arXiv

Other Link： http://arxiv.org/pdf/2011.13588v1

Authorship：Lead author   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

DOI： 10.1109/iv47402.2020.9304681

arXiv

Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

DOI： 10.1109/iv47402.2020.9304809

Scopus

Publishing type：Research paper (scientific journal)   Publisher：Institute of Electrical and Electronics Engineers (IEEE)  

DOI： 10.1109/tvt.2020.2980197

Web of Science

Scopus

Publishing type：Part of collection (book)   Publisher：Springer International Publishing  

DOI： 10.1007/978-3-030-60337-3_3

Scopus

Publishing type：Research paper (scientific journal)   Publisher：Institute of Electrical and Electronics Engineers (IEEE)  

© 2013 IEEE. Automated vehicle technology has recently become reliant on 3D LiDAR sensing for perception tasks such as mapping, localization and object detection. This has led to a rapid growth in the LiDAR manufacturing industry with several competing makers releasing new sensors regularly. With this increased variety of LiDARs, each with different properties such as number of laser emitters, resolution, field-of-view, and price tags, a more in-depth comparison of their characteristics and performance is required. This work compares 10 commonly used 3D LiDARs, establishing several metrics to assess their performance. Various outstanding issues with specific LiDARs were qualitatively identified. The accuracy and precision of individual LiDAR beams and accumulated point clouds are evaluated in a controlled environment at distances from 5 to 180 meters. Reflective targets were used to characterize intensity patterns and quantify the impact of surface reflectivity on accuracy and precision. A vehicle and pedestrian mannequin were also used as additional targets of interest. A thorough assessment of these LiDARs is given with their potential applicability for automated driving tasks. The data collected in these experiments and analysis tools are all shared openly.

DOI： 10.1109/access.2020.3009680

Web of Science

Scopus

Publishing type：Research paper (scientific journal)   Publisher：Fuji Technology Press Ltd.  

Subjective risk assessment is an important technology for enhancing driving safety, because an individual adjusts his/her driving behavior according to his/her own subjective perception of risk. This study presents a novel framework for modeling personalized subjective driving risk during expressway lane changes. The objectives of this study are twofold: (i) to use ego vehicle driving signals and surrounding vehicle locations in a data-driven and explainable approach to identify the possible influential factors of subjective risk while driving and (ii) to predict the specific individual’s subjective risk level just before a lane change. We propose the personalized subjective driving risk model, a combined framework that uses a random forest-based method optimized by genetic algorithms to analyze the influential risk factors, and uses a bidirectional long short term memory to predict subjective risk. The results demonstrate that our framework can extract individual differences of subjective risk factors, and that the identification of individualized risk factors leads to better modeling of personalized subjective driving risk.

DOI： 10.20965/jrm.2020.p0503

Web of Science

Scopus

Publishing type：Part of collection (book)   Publisher：De Gruyter  

DOI： 10.1515/9783110669787-014

Publishing type：Research paper (scientific journal)  

In this work, we present a detailed comparison of ten different 3D LiDAR
sensors, covering a range of manufacturers, models, and laser configurations,
for the tasks of mapping and vehicle localization, using as common reference
the Normal Distributions Transform (NDT) algorithm implemented in the
self-driving open source platform Autoware. LiDAR data used in this study is a
subset of our LiDAR Benchmarking and Reference (LIBRE) dataset, captured
independently from each sensor, from a vehicle driven on public urban roads
multiple times, at different times of the day. In this study, we analyze the
performance and characteristics of each LiDAR for the tasks of (1) 3D mapping
including an assessment map quality based on mean map entropy, and (2) 6-DOF
localization using a ground truth reference map.

arXiv

Other Link： http://arxiv.org/pdf/2004.01374v1

Publishing type：Research paper (scientific journal)   Publisher：Institute of Electrical and Electronics Engineers (IEEE)  

DOI： 10.1109/access.2020.2983149

Web of Science

Scopus

arXiv

Publishing type：Research paper (scientific journal)   Publisher：Institute of Electrical and Electronics Engineers (IEEE)  

DOI： 10.1109/tits.2019.2956066

Web of Science

Scopus

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

Due to multi-path effects, GNSS-based localization methods are not always reliable in urban transportation scenes. To solve this problem, matching-based methods, which compare the real-time sensor data with a prior map, are widely used for urban autonomous driving. In these methods, high-precision noise-free 3D map plays an essential role in vehicle localization. This paper proposes a 3D map optimization framework to generate such map with high efficiency and low memory consumption. First, a deep learning based method is designed to automatically filter out non-map objects in mobile laser scans during mapping. Then, a method named dynamic local NDT is introduced for mapping and localization to improve efficiency and reduce memory usage. Furthermore, a road segmentation method is exploited for further optimization. The proposed framework only relies on LIDAR and GNSS-INS, which makes it simple and easily conducted. The mapping and positioning experimental results show that the proposed framework outperforms the conventional NDT method.

DOI： 10.1109/itsc.2019.8917130

Web of Science

Scopus

Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

© 2019 IEEE. We propose a data-driven control framework for autonomous driving which combines learning-based risk assessment with personalized, safety-focused, predictive control. Different control strategies are used depending on the detected risk level of the driving situation (risky vs. non-risky). This requires a model which can understand the context of the driving situation. In addition, autonomous driving should also be able to provide various safe and comfortable driving styles customized for various users, which requires a modeling method that can capture individual driving preferences. To achieve this, we propose a novel vehicle control framework in which Model Predictive Control (MPC) is combined with a learning-based risk assessment model. Random Forest (RF) methods are trained to classify driving scenes as risky or not risky, while at the same time capturing individually preferred travel velocities. If driving scenes are classified as risky, then the Safety-focused Model Predictive Control (SMPC) system will be launched to generate control commands satisfying predetermined safety constraints, otherwise, Personalized Model Predictive Control (PMPC) is used instead to track the driver's individually preferred velocity. We demonstrate experimentally our control framework.

DOI： 10.1109/itsc.2019.8917457

Web of Science

Scopus

Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

DOI： 10.1109/itsc.2019.8917152

Web of Science

Scopus

Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

DOI： 10.1109/itsc.2019.8917386

Web of Science

Scopus

Publishing type：Research paper (scientific journal)   Publisher：Institute of Electrical and Electronics Engineers (IEEE)  

DOI： 10.1109/access.2019.2935253

Web of Science

Scopus

Publishing type：Research paper (international conference proceedings)  

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

The use of 3D LiDAR, which has proven its capabilities in autonomous driving systems, is now expanding into many other fields. The sharing and transmission of point cloud data from 3D LiDAR sensors has broad application prospects in robotics. However, due to the sparseness and disorderly nature of this data, it is difficult to compress it directly into a very low volume. A potential solution is utilizing raw LiDAR data. We can rearrange the raw data from each frame losslessly in a 2D matrix, making the data compact and orderly. Due to the special structure of 3D LiDAR data, the texture of the 2D matrix is irregular, in contrast to 2D matrices of camera images. In order to compress this raw, 2D formatted LiDAR data efficiently, in this paper we propose a method which uses a recurrent neural network and residual blocks to progressively compress one frame's information from 3D LiDAR. Compared to our previous image compression based method and generic octree point cloud compression method, the proposed approach needs much less volume while giving the same decompression accuracy. Potential application scenarios for point cloud compression are also considered in this paper. We describe how decompressed point cloud data can be used with SLAM (simultaneous localization and mapping) as well as for localization using a given map, illustrating potential uses of the proposed method in real robotics applications.

DOI： 10.1109/icra.2019.8794264

Web of Science

Scopus

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

© 2018 IEEE. End-to-end navigation refers to methods of generating control signals for mobile autonomous devices directly from external sensors, which is an area gaining attention within the autonomous driving research community. Previous autonomous driving research has mostly been focused on keeping moving vehicles within their lanes, but reliable navigation along other trajectories, including branching off onto other routes, has not yet been achieved. In this study we propose a deep learning system for end-to-end navigation which would allow an autonomous vehicle to turn at intersections. Our system's inputs include camera images and directions to a target, while the outputs are the steering control signals needed to direct the vehicle. We validate the system's performance by conducting experiments involving three different driving scenarios: short, indoor trajectories containing a single branching turn; long, outdoor trajectories containing many branching turns; and long, outdoor trajectories which were not included during training. Our end-to-end navigation system allowed an autonomous robot to successfully follow outdoor trajectories with right and left turns, including those which were not part of the training course.

DOI： 10.1109/robio.2018.8665079

Web of Science

Scopus

Authorship：Lead author   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

© 2018 IEEE. In this work we propose a new method to create high density ground maps for autonomous vehicles based on low boundary elevation estimation. Ground maps are created using ray casting of laser beams in 3D LiDAR data accumulated over time and computing the lower elevation data between 3D LiDAR scans. This lower boundary works as a low envelope of the accumulated point cloud. Our lower boundary ground maps approach is not affected by moving objects on the road, and produces high density maps even with coarse vertical resolution 3D LiDARs, works on long curved roads with different elevations, and it is computationally efficient. As proof of concept, we present an application for real-time obstacle and ground segmentation of 3D LiDAR data using our ground maps.

DOI： 10.1109/itsc.2018.8569764

Web of Science

Scopus

Publishing type：Research paper (scientific journal)   Publisher：Fuji Technology Press Ltd.  

Autonomous mobile robot navigation in real unmodified outdoor areas frequented by people on their business, children playing, fast running bicycles, and even robots, remains a difficult challenge. For eleven years, the Tsukuba Challenge Real World Robot Challenge (RWRC) has brought together robots, researchers, companies, government, and ordinary citizens, under the same outdoor space to push forward the limits of autonomous mobile robots. For the Tsukuba Challenge 2017 participation, our team proposed to study the problem of sensors-to-actuators navigation (also called End-to-End), this is, having the robot to navigate towards the destination on a complex path, not only moving straight but also turning at intersections. End-to-End (E2E) navigation was implemented using a convolutional neural network (CNN): the robot learns how to go straight, turn left, and turn right, using camera images and trajectory data. E2E network training and evaluation was performed at Nagoya University, on similar outdoor conditions to that of Tsukuba Challenge 2017 (TC2017). Even thought E2E was trained on a different environment and conditions, the robot successfully followed the designated trajectory in the TC2017 course. Learning how to follow the road no matter the environment is of the key attributes of E2E based navigation. Our E2E does not perform obstacle avoidance and can be affected by illumination and seasonal changes. Therefore, to improve safety and add fault tolerance measures, we developed an E2E navigation approach with model-based system as backup. The model-based system is based on our open source autonomous vehicle software adapted to use on a mobile robot. In this work we describe our approach, implementation, experiences and main contributions.

DOI： 10.20965/jrm.2018.p0563

Web of Science

Scopus

Publishing type：Research paper (scientific journal)   Publisher：Fuji Technology Press Ltd.  

Navigation in social environments, in the absence of traffic rules, is the difficult task at the core of the annual Tsukuba Challenge. In this context, a better understanding of the soft rules that influence social dynamics is key to improve robot navigation. Prior research attempts to model social behavior through microscopic interactions, but the resulting emergent behavior depends heavily on the initial conditions, in particular the macroscopic setting. As such, data-driven studies of pedestrian behavior in a fixed environment may provide key insight into this macroscopic aspect, but appropriate data is scarcely available. To support this stream of research, we release an open-source dataset of dynamic object trajectories localized in a map of 2017 Tsukuba Challenge environment. A data collection platform equipped with lidar, camera, IMU, and odometry repeatedly navigated the challenge’s course, recording observations of passersby. Using a background map, we localized ourselves in the environment, removed the static background from the point cloud data, clustered the remaining points into dynamic objects and tracked their movements over time. In this work, we present the Tsukuba Challenge Dynamic Object Tracks dataset, which features nearly 10,000 trajectories of pedestrians, cyclists, and other dynamic agents, in particular autonomous robots. We provide a 3D map of the environment used as global frame for all trajectories. For each trajectory, we provide at regular time intervals an estimated position, velocity, heading, and rotational velocity, as well as bounding boxes for the objects and segmented lidar point clouds. As additional contribution, we provide a discussion which focuses on some discernible macroscopic patterns in the data.

DOI： 10.20965/jrm.2018.p0598

Web of Science

Scopus

J-GLOBAL

Language：English   Publishing type：Research paper (scientific journal)   Publisher：Springer Science and Business Media LLC  

Reliable people detection is an important task in several areas like security, intelligent environments and human robot interaction. People detection does not depend only upon separation of static environment objects from those showing motion (hopefully humans), a reliable system should be able to detect static people even in cluttered environments.This work presents a reliable approach for people detection and position estimation using multiple layers of Laser Range Finders (LRF) on a mobile robot. Each layer combines two LRF sensors to scan around the robot's surroundings and are vertically separated to detect distinct parts of the human body. By using AdaBoost we create strong classifiers to detect body parts, candidate segments in each layer are fused for people detection, and we use simple data association to estimate their positions. Additionally, this work introduces laser reflection intensity as a novel property for people detection. First, we present a study of laser intensity and textiles, then introduce new intensity-based features for detection, and propose a method for segment separation using laser intensity. We provide a thorough evaluation of our multi-layered system though several experiments on a mobile robot.

DOI： 10.1007/s12369-010-0086-3

Web of Science

Scopus

J-GLOBAL

Other Link： http://link.springer.com/article/10.1007/s12369-010-0086-3/fulltext.html

Authorship：Lead author,　Corresponding author   Language：English   Publishing type：Doctoral thesis  

Authorship：Lead author,　Corresponding author   Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

Effective detection of people is a basic requirement for robot coexistence in human environments. In our previous work [1] we proposed a method for people detection and position estimation using multiple layers of Laser Range Finders (LRF) in a mobile robot. We extend our work by introducing laser reflection intensity as a novel feature for people detection, achieving significant improvement of detection rates. In concrete, we propose a method for calibration of laser intensity data, a method for segment separation using laser intensity, and introduce two new intensity-based features for people detection: the variance of laser intensity and the variance of intensity differences. We present experimental results that confirm the effectiveness of our multi-layered detection method including laser intensity.

DOI： 10.1109/iros.2010.5649769

Web of Science

Scopus

Authorship：Lead author   Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

Successful detection of people is a basic requirement for a robot to achieve symbiosis in people's daily life. Specifically, a mobile robot designed to follow people needs to keep track of people's position through time, for it defines the robot's position and trajectory.In this work we introduce the usage of reflection intensity data of Laser Range Finders (LRF) arranged in multiple layers for people detection. We use supervised learning to train strong classifiers including intensity-based features. Concretely, we propose a calibration method for laser intensity and introduce new intensity-based features for people detection which are combined with range-based features in a strong classifier using supervised learning. We provide experimental results to evaluate the effectiveness of these features.This work is an step towards of our main research project of developing a social autonomous mobile robot acting as member of a people group.

DOI： 10.1109/roman.2010.5598657

Web of Science

Scopus

J-GLOBAL

Authorship：Lead author  

Authorship：Lead author   Publishing type：Part of collection (book)   Publisher：Springer Berlin Heidelberg  

Successful detection and tracking of people is a basic requirement to achieve a robot symbiosis in people daily life. Specifically, a mobile robot designed to follow people needs to keep track of people position through time, for it defines the robot's position and trajectory. This work proposes a new method people detection and position estimation from a mobile robot by fusion of multiple Laser Range Finders arranged in two layers. Sensors facing opposite directions in a single row (layer) are combined to produce a 360° representation of robot's surroundings, then data from every layer is further fused to create a 3D model of people and from there their position. The main problem of our research is an autonomous mobile robot acting as member of a people group moving in public areas, simple but accurate people detection and tracking is an important requirement. We present experimental results of fusion steps and people detection in an indoor environment. © 2009 Springer-Verlag Berlin Heidelberg.

DOI： 10.1007/978-3-540-89859-7_22

Scopus

J-GLOBAL

Language：English   Publishing type：Research paper (scientific journal)   Publisher：WILEY  

This paper describes an implementation of a mobile robot system for autonomous navigation in outdoor concurred walkways. The task was to navigate through nonmodified pedestrian paths with people and bicycles passing by. The robot has multiple redundant sensors, which include wheel encoders, an inertial measurement unit, a differential global positioning system, and four laser scanner sensors. All the computation was done on a single laptop computer. A previously constructed map containing waypoints and landmarks for position correction is given to the robot. The robot system's perception, road extraction, and motion planning are detailed. The system was used and tested in a 1-km autonomous robot navigation challenge held in the City of Tsukuba, Japan, named "Tsukuba Challenge 2007." The proposed approach proved to be robust for outdoor navigation in cluttered and crowded walkways, first on campus paths and then running the challenge course multiple times between trials and the challenge final. The paper reports experimental results and overall performance of the system. Finally the lessons learned are discussed. The main contribution of this work is the report of a system integration approach for autonomous outdoor navigation and its evaluation. © 2009 Wiley Periodicals, Inc.

DOI： 10.1002/rob.20301

Web of Science

Scopus

J-GLOBAL

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

This paper presents and describes the approach for achieving long distance autonomous navigation with a mobile robot on outdoor cluttered pedestrian paths. The task was to finish an event launched by the City of Tsukuba in Japan, called "Real World Robot Challenge", of navigating 1km autonomously in a real environment with real pedestrians and bicycles. The hardware, software and strategy for navigating in cluttered environments is explained. Moreover, the complementary functionality of the overall system where map-based and sensor-based navigation seamlessly change, is presented. The robustness of the system is validated with experimental results.

DOI： 10.1109/iros.2008.4650584

Web of Science

Scopus

Authorship：Lead author   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

DOI： 10.1109/mfi.2008.4648023

Web of Science

Scopus

Authorship：Lead author   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

DOI： 10.1109/mfi.2008.4648075

J-GLOBAL

Authorship：Lead author   Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

Time is crucial in applications such as sensor data fusion, autonomous mobility and SLAM. However clocks at end systems are rarely synchronized and often running at different speeds. Lack of synchronization therefore reduces the accuracy of sensor readings. The SOKUIKI scanning laser range finder allows acquiring time values by timestamping range readings. Our work consists in a method for time synchronization, with clock skew estimation, between a laser range finder sensor and a host computer. ©2008 IEEE.

DOI： 10.1109/MFI.2008.4648075

Web of Science

Scopus

J-GLOBAL

Authorship：Lead author,　Corresponding author   Publishing type：Research paper (international conference proceedings)  

In recent years, Web Caching has been considered one of the key areas to improve web usage efficiency. However, caching web objects proposes many considerations about the validity of the cache. Ideally, it would be valuable to have a consistent cache, where no invalid relationships among objects are held. Several alternatives have been offered to keep consistency in the web cache, each one being better in different situations and for diverse requirements. Usually, web cachers implement just one strategy for maintaining consistency, sometimes giving bad results if circumstances are not appropriate for such strategy. Given that, a web cacher where this policy can be adapted to different situations, will offer good results in an execution with changing conditions. A web caching architecture is proposed as a testbed for consistency models, allowing both timing and ordering issues to be considered.

Scopus