Search

Smooth handling of pedestrian interactions is a key requirement for Autonomous Vehicles (AV) and Advanced Driver Assistance Systems (ADAS). Such systems call for early and accurate prediction of a pedestrian's crossing/not-crossing behaviour in front of the vehicle. Existing approaches to pedestrian behaviour prediction make use of pedestrian motion, his/her location in a scene and static context variables such as traffic lights, zebra crossings etc. We stress on the necessity of early prediction for smooth operation of such systems. We introduce the influence of vehicle interactions on pedestrian intention for this purpose. In this paper, we show a discernible advance in prediction time aided by the inclusion of such vehicle interaction context. We apply our methods to two different datasets, one in-house collected - NTU dataset and another public real-life benchmark - JAAD dataset. We also propose a generic graphical model Factored Latent-Dynamic Conditional Random Fields (FLDCRF) for single and multi-label sequence prediction as well as joint interaction modeling tasks. FLDCRF outperforms Long Short-Term Memory (LSTM) networks across the datasets ($\sim$100 sequences per dataset) over identical time-series features. While the existing best system predicts pedestrian stopping behaviour with 70\% accuracy 0.38 seconds before the actual events, our system achieves such accuracy at least 0.9 seconds on an average before the actual events across datasets., Comment: Accepted by IEEE Transactions on Intelligent Transportation Systems

In this paper, we study the actor-action semantic segmentation problem, which requires joint labeling of both actor and action categories in video frames. One major challenge for this task is that when an actor performs an action, different body parts of the actor provide different types of cues for the action category and may receive inconsistent action labeling when they are labeled independently. To address this issue, we propose an end-to-end region-based actor-action segmentation approach which relies on region masks from an instance segmentation algorithm. Our main novelty is to avoid labeling pixels in a region mask independently - instead we assign a single action label to these pixels to achieve consistent action labeling. When a pixel belongs to multiple region masks, max pooling is applied to resolve labeling conflicts. Our approach uses a two-stream network as the front-end (which learns features capturing both appearance and motion information), and uses two region-based segmentation networks as the back-end (which takes the fused features from the two-stream network as the input and predicts actor-action labeling). Experiments on the A2D dataset demonstrate that both the region-based segmentation strategy and the fused features from the two-stream network contribute to the performance improvements. The proposed approach outperforms the state-of-the-art results by more than 8% in mean class accuracy, and more than 5% in mean class IOU, which validates its effectiveness., Comment: Accepted by BMVC 2018

We consider a single non-holonomic Dubins-like robot traveling with a constant longitudinal speed in an a priori unknown and unsteady planar environment. The robot should detect, locate, and track the boundary of a dynamic environmental scalar field. The field is measured by an on-board sensor in a point-wise fashion at the robot's location. The focus is on unsteady boundaries that evolve over time in an arbitrary fashion, including deformations, i.e., changes of shapes and sizes. We present a sliding mode control method for localizing and tracking such boundaries: the robot is steered to the boundary and circulates in its close proximity afterwards. The proposed control algorithm does not require estimation of the spatial gradient of the field and is non-demanding with respect to both computation and motion. The paper offers the proofs of technical facts required for rigorous justification of non-local convergence of the proposed control law, as well as theoretical illustrations of its performance in specific scenarios., Comment: 3 figures

We consider a single kinematically controlled robot with a bounded control range. The robot travels in a two-dimensional region supporting an unknown unsteady scalar field. A single sensor provides the field value at the current location of the robot. The paper is concerned with a reactive navigation strategy to drive the robot to the time-varying location where the field attains its spatial maximum, with subsequent keeping the robot close to this location. This strategy is based on a novel paradigm of kinematic control different from conventionally trying to align the velocity vector with the field gradient, does not employ estimation of the entire field gradient or derivative-dependent quantities, like the rate at which the available field reading evolves over time, and is non-demanding with respect to both computation and motion. The paper presents the proofs of the technical facts required to justify convergence of the proposed algorithm., Comment: 19 pages, 4 figures

We present a method for guidance of a Dubins-like vehicle with saturated control towards a target in a steady simply connected maze-like environment. The vehicle always has access to to the target relative bearing angle (even if the target is behind the obstacle or is far from the vehicle) and the distance to the nearest point of the maze if it is within the given sensor range. The proposed control law is composed by biologically inspired reflex-level rules. Mathematically rigorous analysis of this law is provided; its convergence and performance are confirmed by computer simulations and experiments with real robots.

Navigation and guidance of autonomous vehicles is a fundamental problem in robotics, which has attracted intensive research in recent decades. This report is mainly concerned with provable collision avoidance of multiple autonomous vehicles operating in unknown cluttered environments, using reactive decentralized navigation laws, where obstacle information is supplied by some sensor system. Recently, robust and decentralized variants of model predictive control based navigation systems have been applied to vehicle navigation problems. Properties such as provable collision avoidance under disturbance and provable convergence to a target have been shown; however these often require significant computational and communicative capabilities, and don't consider sensor constraints, making real time use somewhat difficult. There also seems to be opportunity to develop a better trade-off between tractability, optimality, and robustness. The main contributions of this work are as follows; firstly, the integration of the robust model predictive control concept with reactive navigation strategies based on local path planning, which is applied to both holonomic and unicycle vehicle models subjected to acceleration bounds and disturbance; secondly, the extension of model predictive control type methods to situations where the information about the obstacle is limited to a discrete ray-based sensor model, for which provably safe, convergent boundary following can be shown; and thirdly the development of novel constraints allowing decentralized coordination of multiple vehicles using a robust model predictive control type approach, where a single communication exchange is used per control update, vehicles are allowed to perform planning simultaneously, and coherency objectives are avoided., Comment: 308 pages

We present technical results justifying a method for guidance of a Dubins-like vehicle with saturated control towards a target in a steady simply connected maze-like environment. The vehicle always has access to to the target relative bearing angle and the distance to the nearest point of the maze if it is within the given sensor range. The proposed control law is composed by biologically inspired reflex-level rules. Mathematically rigorous analysis of this law is provided; its convergence and performance are confirmed by computer simulations and experiments with real robots., Comment: 16 pages, 4 figures