This paper studies the problem of detection and tracking of general objects with semistatic dynamics observed by a mobile robot moving in a large environment. A key problem is that due to the environment scale, the robot can only observe a subset of the objects at any given time. Since some time passes between observations of objects in different places, the objects might be moved when the robot is not there. We propose a model for this movement in which the objects typically only move locally, but with some small probability they jump longer distances through what we call global motion. For filtering, we decompose the posterior over local and global movements into two linked processes. The posterior over the global movements and measurement associations is sampled, while we track the local movement analytically using Kalman filters. This novel filter is evaluated on point cloud data gathered autonomously by a mobile robot over an extended period of time. We show that tracking jumping objects is feasible, and that the proposed probabilistic treatment outperforms previous methods when applied to real world data. The key to efficient probabilistic tracking in this scenario is focused sampling of the object posteriors.

In this paper we introduce a system for unsupervised object discovery and segmentation of RGBD-images. The system models the sensor noise directly from data, allowing accurate segmentation without sensor specific hand tuning of measurement noise models making use of the recently introduced Statistical Inlier Estimation (SIE) method. Through a fully probabilistic formulation, the system is able to apply probabilistic inference, enabling reliable segmentation in previously challenging scenarios. In addition, we introduce new methods for filtering out false positives, significantly improving the signal to noise ratio. We show that the system significantly outperform state-of-the-art in on a challenging real-world dataset.

In this paper we introduce an adaptive cost function for pointcloud registration. The algorithm automatically estimates the sensor noise, which is important for generalization across different sensors and environments. Through experiments on real and synthetic data, we show significant improvements in accuracy and robustness over state-of-the-art solutions.

In this paper, a system for incrementally building and maintaining a database of 3D objects for robots with long run times is presented. The system is a step towards lifelong autonomous object modelling using a mobile robot. The proposed solution iteratively fuses observations as they arrive into better and better models. By greedily allowing the system to fuse data, mistakes can be made. The system continuously seek to detect and remove such errors, without the need for batch updates using all known data at once.

We present a novel method for clustering segmented dynamic parts of indoor RGB-D scenes across repeated observations by performing an analysis of their spatial-temporal distributions. We segment areas of interest in the scene using scene differencing for change detection. We extend the Meta-Room method and evaluate the performance on a complex dataset acquired autonomously by a mobile robot over a period of 30 days. We use an initial clustering method to group the segmented parts based on appearance and shape, and we further combine the clusters we obtain by analyzing their spatial-temporal behaviors. We show that using the spatial-temporal information further increases the matching accuracy.

The video presents a novel approach for vision-based topological localisation in dynamic indoor environments. In contrast to other approaches that rely on static image features, our method explicitly models the temporal dynamics of the visibility of image features. The proposed spatio-temporal world model is able to predict the visibility of image features at different times of day, which allows to construct time-specific environment appearance models. This approach improves the robot localisation capabilities during long-term operations.

In this work we present the Probabilistic Primitive Refinement (PPR) algorithm, an iterative method for accurately determining the inliers of an estimated primitive (such as planes and spheres) parametrization in an unorganized, noisy point cloud. The measurement noise of the points belonging to the proposed primitive surface are modelled using a Gaussian distribution and the measurements of extraneous points to the proposed surface are modelled as a histogram. Given these models, the probability that a measurement originated from the proposed surface model can be computed. Our novel technique to model the noisy surface from the measurement data does not require a priori given parameters for the sensor noise model. The absence of sensitive parameters selection is a strength of our method. Using the geometric information obtained from such an estimate the algorithm then builds a color-based model for the surface, further boosting the accuracy of the segmentation. If used iteratively the PPR algorithm can be seen as a variation of the popular mean-shift algorithm with an adaptive stochastic kernel function.

This paper presents a new approach for topological localisation of service robots in dynamic indoor environments. In contrast to typical localisation approaches that rely mainly on static parts of the environment, our approach makes explicit use of information about changes by learning and modelling the spatio-temporal dynamics of the environment where the robot is acting. The proposed spatio-temporal world model is able to predict environmental changes in time, allowing the robot to improve its localisation capabilities during long-term operations in populated environments. To investigate the proposed approach, we have enabled a mobile robot to autonomously patrol a populated environment over a period of one week while building the proposed model representation. We demonstrate that the experience learned during one week is applicable for topological localization even after a hiatus of three months by showing that the localization error rate is significantly lower compared to static environment representations.

Finding accurate correspondences between overlapping 3D views is crucial for many robotic applications, from multi-view 3D object recognition to SLAM. This step, often referred to as view registration, plays a key role in determining the overall system performance. In this paper, we propose a fast and simple method for registering RGB-D data, building on the principle of the Iterative Closest Point (ICP) algorithm. In contrast to ICP, our method exploits both point position and visual appearance and is able to smoothly transition the weighting between them with an adaptive metric. This results in robust initial registration based on appearance and accurate final registration using 3D points. Using keypoint clustering we are able to utilize a non exhaustive search strategy, reducing runtime of the algorithm significantly. We show through an evaluation on an established benchmark that the method significantly outperforms current methods in both robustness and precision.