Abstract—In this paper, we present an approach for modeling 3D environments based on octrees using a probabilistic occupancy estimation. Our technique is able to represent full 3D models including free and unknown areas. It is available as an open-source library to facilitate the development of 3D mapping systems. We also provide a detailed review of existing approaches to 3D modeling. Our approach was thoroughly evaluated using different real-world and simulated datasets. The results demonstrate that our approach is able to model the data probabilistically while, at the same time, keeping the memory requirement at a minimum. I.

Abstract—Mobile robots that interact with humans in an intuitive way must be able to follow directions provided by humans in unconstrained natural language. In this work we investigate how statistical machine translation techniques can be used to bridge the gap between natural language route instructions and a map of an environment built by a robot. Our approach uses training data to learn to translate from natural language instructions to an automatically-labeled map. The complexity of the translation process is controlled by taking advantage of physical constraints imposed by the map. As a result, our technique can efficiently handle uncertainty in both map labeling and parsing. Our experiments demonstrate the promising capabilities achieved by our approach. Index Terms—Human-robot interaction; instruction following; navigation; statistical machine translation; natural language I.

Abstract — We report our first experiences with Leaving Flatland, an exploratory project that studies the key challenges of closing the loop between autonomous perception and action on challenging terrain. We propose a comprehensive system for localization, mapping, and planning for the RHex mobile robot in fully 3D indoor and outdoor environments. This system integrates Visual Odometry-based localization with new techniques in real-time 3D mapping from stereo data. The motion planner uses a new decomposition approach to adapt existing 2D planning techniques to operate in 3D terrain. We test the map-building and motion-planning subsystems on real and synthetic data, and show that they have favorable computational performance for use in high-speed autonomous navigation. I.

Abstract — In this paper, we present a localization method for humanoid robots navigating in arbitrary complex indoor environments using only onboard sensing. Reliable and accurate localization for humanoid robots operating in such environments is a challenging task. First, humanoids typically execute motion commands rather inaccurately and odometry can be estimated only very roughly. Second, the observations of the small and lightweight sensors of most humanoids are seriously affected by noise. Third, since most humanoids walk with a swaying motion and can freely move in the environment, e.g., they are not forced to walk on flat ground only, a 6D torso pose has to be estimated. We apply Monte Carlo localization to globally determine and track a humanoid’s 6D pose in a 3D world model, which may contain multiple levels connected by staircases. To achieve a robust localization while walking and climbing stairs, we integrate 2D laser range measurements as well as attitude data and information from the joint encoders. We present simulated as well as real-world experiments with our humanoid and thoroughly evaluate our approach. As the experiments illustrate, the robot is able to globally localize itself and accurately track its 6D pose over time. I.

Abstract — This paper presents a visual servoing scheme to control humanoid dynamic walk. Whereas most of the existing approaches follow a perception-decision-action scheme, we hereby introduce a method that uses the on-line information given by an on-board camera. This close looped approach allows the system to react to changes in its environment and adapt to modelling error. Our approach is based on a new reactive pattern generator which modifies footsteps, center of mass and center of pressure trajectories at the control level for the center of mass to track a reference velocity. In this workshop, we present three ways of servoing dynamical humanoïd walk: a naïve one that compute a reference velocity using a visual servoing control law, a second one that takes into account the sway motion induced by the walk and an on going work on vision predictive control that directly introduces the visual error in the cost function of the pattern generator. The two first approaches have been validated on the HRP-2 robot. These close loop approaches give a more accurate positioning than the one obtained when executing a planned trajectory especially when rotational motion are involved. I.

Abstract—This paper presents a modular software architecture based on a stereo camera system that makes real-time 3D perception, interaction and navigation for small humanoid robots possible. The hardware-independent architecture can be used for several purposes, like 3D visualisation, object recognition, selflocalization, collision avoidance et cetera. First of all, the intrinsic calibration parameters of the stereo camera system are identified. Then the lens distortions of the input images are revised, and both images are rectified according to the extrinsic parameters to facilitate the subsequent correspondence analysis. The imaged scenery, as well as a region of interest, can be reconstructed either by the block-matching, the Schirai- or the Birchfield algorithm, followed by triangulation back into 3D space. Finally, the above-mentioned aims can be tackled based on the resulting 3D model computed by our modular software architecture. Index Terms—software architecture, real-time, perception of environment, stereo vision, 3D reconstruction, humanoid robot. I.

Abstract — In this paper, we present an approach to obstacle detection for collision-free, efficient humanoid robot navigation based on monocular images and sparse laser range data. To detect arbitrary obstacles in the surroundings of the robot, we analyze 3D data points obtained from a 2D laser range finder installed in the robot’s head. Relying only on this laser data, however, can be problematic. While walking, the floor close to the robot’s feet is not observable by the laser sensor, which inherently increases the risk of collisions, especially in nonstatic scenes. Furthermore, it is time-consuming to frequently stop walking and tilting the head to obtain reliable information about close obstacles. We therefore present a technique to train obstacle detectors for images obtained from a monocular camera also located in the robot’s head. The training is done online based on sparse laser data in a self-supervised fashion. Our approach projects the obstacles identified from the laser data into the camera image and learns a classifier that considers color and texture information. While the robot is walking, it then applies the learned classifiers to the images to decide which areas are traversable. As we illustrate in experiments with a real humanoid, our approach enables the robot to reliably avoid obstacles during navigation. Furthermore, the results show that our technique leads to significantly more efficient navigation compared to extracting obstacles solely based on 3D laser range data acquired while the robot is standing at certain intervals. I.