2016 IEEE Intelligent Transportation Systems Conference

8th Workshop on Planning, Perception and Navigation for Intelligent Vehicles
A bridge between Robotics and ITS technologies

Full Day Workshop,

November 1st, 2016, Rio de Janeiro, Brazil

Contact : Research Director Christian Laugier
INRIA, Emotion project, INRIA Rhône-Alpes,
655 Avenue de l'Europe, 38334 Saint Ismier Cedex, France,
Phone: +33 4 7661 5222, Fax : +33 4 7661 5477,
Email: Christian.Laugier@inrialpes.fr,
Home page: http://emotion.inrialpes.fr/laugier



Organizers

Research Director Christian Laugier, INRIA, Emotion project, INRIA Rhône-Alpes, 655 Avenue de l'Europe, 38334 Saint Ismier Cedex, France, Phone: +33 4 7661 5222, Fax : +33 4 7661 5477, Email: Christian.Laugier@inrialpes.fr,
Home page: http://emotion.inrialpes.fr/laugier

Professor Philippe Martinet, IRCCyN-CNRS Laboratory, Ecole Centrale de Nantes, 1 rue de la Noë, 44321 Nantes Cedex 03, France, Phone: +33 237406975, Fax: +33 237406930, Email: Philippe.Martinet@irccyn.ec-nantes.fr,
Home page: http://www.irccyn.ec-nantes.fr/~martinet

Professor Urbano Nunes, Department of Electrical and Computer Engineering of the Faculty of Sciences and Technology of University of Coimbra, 3030-290 Coimbra, Portugal, GABINETE 3A.10, Phone: +351 239 796 287, Fax: +351 239 406 672, Email: urbano@deec.uc.pt,
Home page: http://www.isr.uc.pt/~urbano

Professor Christoph Stiller, , Institut für Mess- und Regelungstechnik, Karlsruher Institut für Technologie (KIT), Engler-Bunte-Ring 21, Gebäude: 40.32, 76131 Karlsruhe, Germany, Phone: +49 721 608-42325 Fax: +49 721 661874, Email: stiller@kit edu
Home page: http://www.mrt.kit.edu/mitarbeiter_stiller.php

General Scope

The purpose of this workshop is to discuss topics related to the challenging problems of autonomous navigation and of driving assistance in open and dynamic environments possibly populated by human beings. Technologies related to application fields such as unmanned outdoor vehicles or intelligent road vehicles will be considered from both the theoretical and technological point of views. Several research questions located on the cutting edge of the state of the art will be addressed. Among the many application areas that robotics is addressing, transportation of people and goods seem to be a domain that will dramatically benefit from intelligent automation. Fully automatic driving is emerging as the approach to dramatically improve efficiency while at the same time leading to the goal of zero fatalities. This workshop will address robotics technologies, which are at the very core of this major shift in the automobile paradigm. Technologies related to this area, such as autonomous outdoor vehicles, achievements, challenges and open questions would be presented and discussed.

Main Topics

  • Perception: Lane detection & lane keeping, Feature extraction & selection, Pedestrian & vehicle detection, Moving objects Detection & Tracking, Objects classification, Real-time perception and sensor fusion, etc.
  • Road scene understanding: Environment perception & understanding, Vehicle localization for autonomous navigation, SLAM in dynamic environments, Mapping & Semantic maps for navigation, Robust sensor-based 3D reconstruction, Prediction techniques, etc.
  • Planning & Decision-making for navigation: Real-time motion planning in dynamic environments, Advanced driver assistance systems, Autonomous navigation, Cooperative navigation & perception techniques, Behavior modeling & learning, Modeling & Control of mobile robots and vehicles, Collision prediction & avoidance, Human-Robot Interaction, etc
  • International Program Committee
    The selection process will be done through the same procedure than ITSC16 conference (peer review process using paperplazza) managed by dedicated Associated Editors.
  • Javier Ibanez-Guzman (Renault, France)
  • Christian Laugier (Emotion, INRIA, France)
  • Philippe Martinet (IRCCyN, Ecole Centrale de Nantes, France)
  • Fawzi Nashashibi (INRIA, France)
  • Urbano Nunes (Coimbra University, Portugal),
  • Anya Petrovskaya (Stanford, USA)
  • Patrick Rives(INRIA, France)
  • Christoph Stiller (Karlsruhe Institute of Technology, Germany)
  • Rafael Toledo Moreo (Universidad Politécnica de Cartagena, Spain)
  • Danwei Wang (NTU, Singapore)
  • Ming Yang (SJTU, China)
  • J. Marius Zöllner (FZI, Germany)
  • Final program

    Introduction to the workshop 8:40-8-50 C. Laugier and P. Martinet

    Session I: Perception & Situation Awareness; 8:50

    • Title: On the use of maps for Automated Driving 8:50
      Keynote speaker: C. Stiller (KIT, Germany)

      Abstract: Vehicle automation is among the most fascinating trends in automotive electronics. We investigate the information needed by automated vehicles. The augmentation of sensor information by prior knowledge from digital maps is elaborated. We show that cognitive and autonomous vehicles with a few close-to-market sensors are feasible when prior information is available from maps. Vision plays the dominant role in our autonomous vehicle. We completely avoid bulky on-roof mounted sensors. The sensor suite enables the vehicle to perceive its environment and automatically navigate through everyday's traffic. Automated mapping as well as real-time automated decision-making and trajectory planning methods are outlined. Extensive experiments are shown in real world scenarios from our AnnieWAY vehicle, the winner of the 2011 and second winner of the 2016 Grand Cooperative Driving Challenge, and from the Bertha vehicle that drove autonomously on the 104 km of the Bertha Benz memorial route from Mannheim to Pforzheim through a highly populated area of Germany.

    • Title: Increasing the Convergence Domain of RGB-D Direct Registration Methods for Vision-based Localization in Large Scale Environments 9:30
      Authors: R. Martins, P. Rives

      Abstract: Developing autonomous vehicles capable of dealing with complex and dynamic unstructured environments over large-scale distances, remains a challenging goal. One of the major difficulties in this objective is the precise localisation of the vehicle within its environment so that autonomous navigation techniques can be employed. In this context, this paper presents a methodology to map building and to efficient pose computation which is specially adapted for cases of large displacements. Our method uses hybrid robust RGBD cost functions that have different convergence properties, whilst exploiting the visibility rotation invariance given by panoramic images (spherical). The proposed registration model is composed of a RGB and point-to-plane ICP cost in a multi-resolution framework. The convergence analysis of this formulation is performed using simulated spherical images from the Sponza Atrium model benchmark. We close up the paper presenting mapping and localization results in simulated and real outdoor scenes.

    • Title: 3D Object Tracking in Driving Environment: a short review and a benchmark dataset 9:50
      Authors: P. Girão, A. Asvadi, P. Peixoto, U. Nunes

      Abstract: Research in autonomous driving has matured significantly with major automotive companies making important efforts to have their autonomous or highly automated cars available in the near future. As driverless cars move from laboratories to complex real-world environments, the perception capabilities of these systems to acquire, model and interpret the 3D spatial information must become more powerful. Object tracking is one of the challenging problems of autonomous driving in 3D dynamic environments. Although different approaches are proposed for object tracking with demonstrated success in driving environments, it is very difficult to evaluate and compare them because they are defined with various constraints and boundary conditions. The appearance modeling for object tracking in the driving environments, using a multimodal perception system of autonomous cars and advanced driver assistance systems (ADASs), and the evaluation of such object trackers are the research focus of this paper. A benchmark dataset, called 3D Object Tracking in Driving Environments (3D-OTD), is also proposed to facilitate the assessment of object appearance modeling in object tracking methods.

    • Title: The Development of an Ontology for Context Modelling for Driving Context Modelling and Reasoning 10:10
      Authors: Z. Xiong, V. Dixit, S. Travis Waller

      Abstract: In order to use technology to influence human behaviour and promote safer and more fuel efficient behaviour through incentive mechanisms, an instrumented vehicle is developed. This study will contribute towards a broader framework for Intelligent Transport Systems that involve autonomous vehicles. As a result, this vehicle is being developed towards full autonomy. The first step is to make it “perceive” the outside world, so extracting knowledge from some data sources such as sensors is crucial. More critically, there is a fundamental need for a standard that would enable knowledge sharing/exchanging among the different entities, e.g., between on-board sensors, invehicle controls and traffic management agencies. This paper proposes an Ontology for Context Modelling (OCM) to be used as the world model for driving context representation and reasoning, which can enable a better understanding of traffic context and sensor capability, which is the basis for providing data source to Advanced Driver Assistance Systems (ADAS), V2X (Vehicle-to-Everything) communications and even driving decision making within autonomous vehicles. Through the experiments, we evaluate the capability of the OCM to represent the driving context and the reasoning mechanism to compensate for sensor failures and recognize lane changing and overtaking events. This methodology has significant value for creating standards in autonomous and semi-autonomous cars.

    Coffee Break 10:30-11:00

    Session II: Mapping, Localization and Dynamic Scenes simulation 11:00
    • Title: Real Time Pedestrian Tracking, Prediction, and Navigation 11:00
      Keynote speaker: Dinesh Manocha

      Abstract: As autonomous robots and vehicles get widely deployed, one key challenge is to develop robust technologies that can automatically detect and safely navigate through moving pedestrians. While humans are naturally equipped to observe and predict the trajectories of other pedestrians and crowds, there are still many open challenges for current autonomous systems. In this talk, I give an overview of our work of developing novel algorithms and systems for automatic pedestrian identification, tracking, trajectory prediction, and safe navigation. We use a combination of off-line deep learning models and online pedestrian dynamics and combing them techniques from computer vision and physics-based simulation. We have evaluated their performance in sparse as well as dense pedestrians videos and observe considerable improvement in accuracy and performance. Our approach can perform all these computations in realtime in streaming videos and we highlighted their performance on real-world benchmarks as well as simulated scenarios.

    • Title: Occupancy Grid based Urban Localization Using Weighted Point Cloud 11:40
      Authors: L. Guo, M. Yang, B. Wang, C. Wang

      Abstract: Localization is considered as a key capability for autonomous vehicles act in urban environments. Though have been proved to be able to preform convictive results, localization methods using neither laser scanners nor vision sensors could achieve the goal about balancing between accuracy and cost. In this paper, an occupancy grid based localization framework is presented in order to obtain a precise positioning result with relatively low-cost sensor configuration in large scale urban environment. The proposed approach takes a prebuilt global grid map as prior knowledge for localization. Model based feature extraction method is introduced to provide laser points classification, with each extracted point allocated a specified weight to describe local characteristic. The prior grid map is generated from weighted point cloud to be able to describe the local metric features such as curbs and building facades. Localization function is then carried out with a weight point based maximum likelihood matching method to determine the correspondence between local point cloud and the reference grid map. There are also position initialization and reference map management modules to make the framework more practical and reliable. In the end, the proposed approach has been validated by promising experimental results with long distance tests in large urban environments.

    • Title: Cooperative Localization via DSRC and Multi-Sensor Multi-Target Track Association 12:00
      Authors: A. Hamdi Sakr, G. Bansal

      Abstract: Vehicles in the near future will be equipped with dedicated short-range communications (DSRC) transceiver which holds great promise of significantly reducing vehicle collisions by enabling vehicle-to-vehicle (V2V) and vehicleto- infrastructure (V2I) communications. In addition, modern vehicles will be equipped with different on-board sensors such as GPS receivers and ranging sensors (e.g., cameras, radars, and lidars). Using these technologies, this paper proposes a comprehensive system design to improve the positioning of a host vehicle based on Kalman filters. In this approach, the host vehicle fuses its own position information obtained by the on-board GPS receiver with position information of nearby vehicles collected by the on-board ranging sensor(s) and the messages received via the DSRC transceiver from other equipped vehicles. This process also involves performing track matching using a multi-sensor multi-target track association algorithm.We provide insights on the system design and present simulation results that show significant performance gains of the proposed method in terms of localization accuracy and matching accuracy.

    Lunch break 12:20-14:00

    Session III: Behaviors Modeling and Learning, Motion Prediction and Decision-Making 14:00
    • Title: Deep Learning for Robot Perception and Navigation 14:00
      Keynote speaker: Wolfram Burgard (Autonomous Mobile System Laboratory, University of Freiburg, Germany)

      Abstract: Autonomous robots are faced with a series of learning problems to optimize their behavior. In this presentation I will describe recent approaches developed in my group based on deep learning architectures for perception and navigation. In particular, I will discuss approaches to object recognition, body part segmentation from RGB(-D) images, terrain classification, and mobile robot navigation. For all approaches I will describe the underlying network architectures as well as the required data augmentation techniques. I will present the results of expensive experiments quantifying in which way the corresponding algorithm extends the state of the art.

    • Title: High-speed highway scene prediction based on driver models obtained from demonstrations 14:40
      Authors: D. Sierra González, J. Steeve Dibangoye, C. Laugier

      Abstract: One of the key factors to ensure the safe operation of autonomous and semi-autonomous vehicles in dynamic environments is the ability to accurately predict the motion of the dynamic obstacles in the scene. In this work, we show how to use a realistic driver model obtained from demonstrations via Inverse Reinforcement Learning to predict the long-term evolution of highway traffic scenes. The interactions between traffic participants are explicitly considered in our framework, while keeping the computational complexity linear in the number of vehicles in the scene. Preliminary experiments in simulated and real scenarios show the capability of our approach to produce reliable and sensible scene predictions.

    • Title: Space Discretization Using Gaussian Process for Road Intersection 15:00
      Authors: M. Barbier, C. Laugier, J. Ibanez Guzman, O. Simonin

      Abstract: A framework to discretize the space within and around a cross intersection is proposed in this paper. The purpose is to capture the manner in which drivers manoeuvre in an intersection in order to facilitate and understand the decision-making tasks. Gaussian processes are used to learn and predict the most likely trajectories taken by multiple drivers in different situations. The merging and crossing areas are found by searching for the overlap between two predicted trajectories, whereas the area approaching the intersection is discretized by using the most probable occupancy. The generated areas are stored in a map. It was possible to show the correlation between this discretization and the drivers’ behaviour by looking at how the proposed framework also discretizes the velocity profile, which can then be applied to decision making.

    Coffee Break 15:40-16:00

    Session IV: Planning and Navigation 16:00
    • Title: Sensor based Navigation 16:00
      Keynote speaker: Philippe Martinet (IRCCyN, Ecole Centrale de Nantes, France)
      Co-authors: Gaetan Garcia (IRCCyN, Ecole Centrale de Nantes, France), Salvador Dominguez Quijada (IRCCyN, Ecole Centrale de Nantes, France)

      Abstract: It is known that 3D based autonomous navigation requires accurate and reliable maps and the use of accurate and robust localization techniques. 3D accurate sensors are costly and for some of them, doesnt insure the same reliability everywhere causing propagation of uncertainties. Sensor based mapping is based on the general technique of teaching by showing which is very well known in Visual servoing community. It doesn't require strong assumption on global mapping as long as a topological level is used. Sensor based topological navigation is using a sensor based mapping technique in order to replay what have been learnt before. It allows to avoid all accumulation of errors (during mapping and/or global localization), and is using local measurements. We will illustrate some applications done in the field of autonomous navigation in inner cities.

    • Title: Optimal Trajectory Planning for Autonomous Driving Integrating Logical Constraints: a MIQP perspective 16:40
      Authors: X. QIAN, F. Altché, P. Bender, C. Stiller, A. de La Fortelle

      Abstract: This paper considers the problem of optimal trajectory generation for autonomous driving under both continuous and logical constraints. Classical approaches based on continuous optimization formulate the trajectory generation problem as a nonlinear program, in which vehicle dynamics and obstacle avoidance requirements are enforced as nonlinear equality and inequality constraints. In general, gradientbased optimization methods are then used to find the optimal trajectory. However, these methods are ill-suited for logical constraints such as those raised by traffic rules, presence of obstacles and, more generally, to the existence of multiple maneuver variants. We propose a new formulation of the trajectory planning problem as a Mixed-Integer Quadratic Program. This formulation can be solved effectively using widely available solvers, and the resulting trajectory is guaranteed to be globally optimal. We apply our framework to several scenarios that are still widely considered as challenging for autonomous driving, such as obstacle avoidance with multiple maneuver choices, overtaking with oncoming traffic or optimal lane-change decision making. Simulation results demonstrate the effectiveness of our approach and its real-time applicability.

    • Title: Autonomous Personal Mobility Scooter for Multi-Class Mobility-on-Demand Service 17:00
      Authors: H. Andersen, Y. Hong Eng, W.K. Leong, C. Zhang, H.X. Kong, S. Pendleton, M. H. Ang Jr., D. Rus

      Abstract: In this paper, we describe the design and development of an autonomous personal mobility scooter that was used in public trials during the 2016 MIT Open House, for the purpose of raising public awareness and interest about autonomous vehicles. The scooter is intended to work cooperatively with other classes of autonomous vehicles such as road cars and golf cars to improve the efficacy of Mobilityon- Demand transportation solutions. The scooter is designed to be robust, reliable, and safe, while operating under prolonged durations. The flexibility in fleet expansion is shown by replicating the system architecture and sensor package that has been previously implemented in the road car and golf cars. We show that the vehicle performed robustly with small localization variance. A survey of the users shows that the public is very receptive to the concept of the autonomous personal mobility device.

    • Title: Autonomous parking using a sensor based approach 17:20
      Authors: D. Pérez-Morales, S. Dominguez Quijada, O. Kermorgant, P. Martinet

      Abstract: This paper considers the perpendicular parking problem of car-like vehicles for both forward and reverse maneuvers. A sensor based controller with a weighted control scheme is proposed and is compared with a state of the art path planning approach. The perception problem is threated as well considering a Velodyne VLP-16 as the sensor providing the required exteroceptive information. A methodology to extract the necessary information for both approaches from the sensor data is presented. A fast prototyping environment has been used to develop the parking application, and also used as simulator in order to validate the approach. Preliminary results show the effectiveness of the proposed approach.

    Closing 17:40-17:50
    Author Information

      Format of the paper: Papers should be prepared according to the ITSC16 final camera ready format and should be 6 pages long. You must use the code 95vpj of the workshop when you submit your paper on paperplazza. The detailed information on the paper format is available from the ITSC16 page. https://web.fe.up.pt/~ieeeitsc2016/?page_id=98.

      Talk information

      • Invited talk: 40 min (35 min talk, 5 min question)
      • Other talk: 20 min (17 min talk, 3 min question)

      Interactive session

      • Interactive and open session: 1h00

    Previous workshops

      Previously, several workshops were organized in the near same field. The 1st edition PPNIV'07 of this workshop was held in Roma during ICRA'07 (around 60 attendees), the second PPNIV'08 was in Nice during IROS'08 (more than 90 registered people), the third PPNIV'09 was in Saint-Louis (around 70 attendees) during IROS'09, the fourth edition PPNIV'12 was in Vilamoura (over 95 attendees) during IROS'12, the fifth edition PPNIV'13 was in Tokyo (over 135 attendees) during IROS'13, the sixth edition PPNIV'14 was in Chicago (over 100 attendees) during IROS'14, and the seventh edition PPNIV'15 was in Hamburg (over 155 attendees) during IROS'15.
      In parallel, we have also organized SNODE'07 in San Diego during IROS'07 (around 80 attendees), SNODE'09 in Kobe during ICRA'09 (around 70 attendees), RITS'10 in Anchrorage during ICRA'10 (around 35 attendees), and the last one PNAVHE11 in San Francisco during the last IROS11 (around 50 attendees), and MEPC'14 was in Hong-Kong during ICRA'14 (over 60 attendees).

      Special issues have been published in IEEE Transaction on ITS (Car and ITS applications, September 2009), in IEEE-RAS Magazine (Perception and Navigation for Autonomous Vehicles, March 2014) and in ITS Magazine (Perception and Navigation for Autonomous Vehicles, March 2016). We also plan to prepare a special issue in the IEEE Transactions on IV.

    Keynotes

      Proceedings: The workshop proceedings will be published within the ITSC16 Proceedings.

      Special issue: Selected papers will be considered for a special issue in the IEEE Transactions on IV . We will issue an open call, submissions will go through a separate peer review process.