The time complexity of making observations and loop closures in a graph-based visual SLAM system is a function of the number of views stored [1], [2]. Clever algorithms, such as approximate nearest neighbor search, can make this function sub-linear. Despite this, over time the number of views can still grow to a point at which the speed and/or accuracy of the system becomes unacceptable, especially in computation- and memory-constrained SLAM systems. However, not all views are created equal. Some views are rarely observed, because they have been created in an unusual lighting condition, or from low quality images, or in a location whose appearance has changed. These views can be removed to improve the overall performance of a SLAM system. In this paper, we propose a method for pruning views in a visual SLAM system to maintain its speed and accuracy for long term use.

We present a fast nonlinear approximation method for marginalizing out nodes on pose graphs for longterm simultaneous localization, mapping, and navigation. Our approximation preserves the pose graph structure to leverage the rich literature of pose graphs and optimization schemes. By re-parameterizing from absolute-to relative-pose spaces, our method does not suffer from the choice of linearization points as in previous works. We then join our approximation process with a scaled version of the recently-demoted pose-composition approach. Our approach eschews the expenses of many state-of-the-art convex optimization schemes through our efficient and simple O(N2) implementation for a given known topology of the approximate subgraph. We demonstrate its speed and near optimality in practice by comparing against state-of-the-art techniques on popular datasets.

Occupancy mapping is an important component of a robot navigation system since it enables intelligent planning of actions to accomplish a task. In some cases, occupancy information is mapped as part of a SLAM system. In other cases, such as visual SLAM, an occupancy map must be separately created, harnessing localization information. We present a novel occupancy mapping method that adaptively and efficiently takes advantage of improved trajectory estimates from the localization system, without requiring storage of all poses of the trajectory or the sensor information from all poses. Occupancy information is stored in local maps anchored to poses in the trajectory. Local maps are created adaptively depending on the uncertainty of localization. The local maps maintain consistency of occupancy information independently of changes in the localization map due to loop closures or convergence of the estimation algorithm. At any time, a global occupancy map (snapshot map) can be rendered from the local occupancy maps. We evaluate the performance of the system using real data and compare it to two baseline methods.

We address the problem of systematically covering all accessible floor space in an unknown environment by a mobile robot. Our approach uses rectangular regions that are swept across the environment. In the first stage, the robot covers each region using the classic boustrophedon pattern and planning paths to uncovered areas within the region while keeping track of its position uncertainty. The region is then moved sideways to cover the next part of the environment until all accessible space has been visited. In the second stage, the robot revisits the perimeter around the obstacles. We compare our method in terms of total trajectory length to 5 off-line methods including the distance transformation by Zelinsky et al. [1] in a standard test environment as well as in multi-room homes. The presented method has been employed in our Mint cleaning robot [2] for autonomously sweeping and mopping the floors.

Localization using continuous signals such as WiFi or active beacons is a cost-effective approach for enabling systematic navigation of robots. In our previous work we showed how localization maps, represented as regular grids, of such signals can be learned through application of vector field SLAM [1]. In this paper we describe a method that, given such a localization map, finds the pose of a mobile robot from observations of the signals. Our method first generates pose hypotheses by searching the localization map for places that best fit to a measurement taken by the robot. A localization filter using an extended Kalman filter (EKF) then verifies one pose hypothesis by tracking the pose over a short distance. In experiments carried out in a standard test environment equipped with active beacons we obtain an average position accuracy of 10 to 35 cm with a localization success rate of 96 to 99 %. The proposed method enables a robot mapping an environment using vector field SLAM to recover from kidnapping and resume its navigation.

We present a novel algorithm, Compact Kd-Trees (CompactKdt), that achieves state-of-the-art performance in searching large scale object image collections. The algorithm uses an order of magnitude less storage and computations by making use of both the full local features (e.g. SIFT) and their compact binary signatures to build and search the K-Tree. We compare classical PCA dimensionality reduction to three methods for generating compact binary representations for the features: Spectral Hashing, Locality Sensitive Hashing, and Locality Sensitive Binary Codes. CompactKdt achieves significant performance gain over using the binary signatures alone, and comparable performance to using the full features alone. Finally, our experiments show significantly better performance than the state-of-the-art Bag of Words (BoW) methods with equivalent or less storage and computational cost.

Localization in unknown environments using low-cost sensors on embedded hardware is challenging. Yet, it is a requirement for consumer robots if systematic navigation is desired. In this paper, we present a localization approach that learns the spatial variation of an observed continuous signal over the environment. We model the signal as a piecewise linear function and estimate its parameters using a simultaneous localization and mapping (SLAM) approach. By applying the concepts of the exactly sparse extended information filter (ESEIF) , a constant-time, linear-space algorithm is obtained under certain approximations. We apply our framework to a sensor measuring bearing to active beacons, where measurements are distorted because of occlusion and signal reflections. Experimental results from running GraphSLAM, extended Kalman filter SLAM, and ESEIF-SLAM on manually collected sensor measurements, as well as on data recorded on a vacuum-cleaner robot, validate our model. The ESEIF-SLAM solution is evaluated on an ARM 7 embedded board with 64-kB RAM connected to a Roomba 510 vacuum cleaner. The presented methods are also used in Evolution Robotics ' Mint Cleaner product for autonomous floor cleaning.

Distributed Kd-Trees is a method for building image retrieval systems that can handle hundreds of millions of images. It is based on dividing the Kd-Tree into a “root subtree” that resides on a root machine, and several “leaf subtrees”, each residing on a leaf machine. The root machine handles incoming queries and farms out feature matching to an appropriate small subset of the leaf machines. Our implementation employs the MapReduce architecture to efficiently build and distribute the Kd-Tree for millions of images. It can run on thousands of machines, and provides orders of magnitude more throughput than the state-of-the-art, with better recognition performance. We show experiments with up to 100 million images running on 2048 machines, with run time of a fraction of a second for each query image.

This year's conference focuses on human-robot interaction in the real world. The panel discussion presents the view from those who are "living it": industry leaders who are relying on current robotic technology to accomplish their work right now. Who better to provide the most compelling information on issues/challenges that are influencing their use? This panel gathers experts from business and industry to discuss their experiences in using robotic technology in their field. Topics include the financial, organizational, and practical challenges faced by professionals using robotic technology in the workplace, and factors influencing the acceptance of robots at work. The implications for design of robotic products and systems are also discussed. The session is a must for professionals and academic researchers interested in solving problems related to using robots in real world settings.

Indexing quickly and accurately in a large collection of images has become an important problem with many applications. Given a query image, the goal is to retrieve matching images in the collection. We compare the structure and properties of seven different methods based on the two leading approaches: voting from matching of local descriptors vs. matching histograms of visual words, including some new methods. We derive theoretical estimates of how the memory and computational cost scale with the number of images in the database. We evaluate these properties empirically on four real-world datasets with different statistics. We discuss the pros and cons of the different methods and suggest promising directions for future research.

Localization in unknown environments using low-cost sensors remains a challenge. This paper presents a new localization approach that learns the spatial variation of an observed continuous signal. We model the signal as a piece-wise linear function and estimate its parameters using a simultaneous localization and mapping (SLAM) approach. We apply our framework to a sensor measuring bearing to active beacons where measurements are systematically distorted due to occlusion and signal reflections of walls and other objects present in the environment. Experimental results from running GraphSLAM and EKF-SLAM on manually collected sensor measurements as well as on data recorded on a vacuum-cleaner robot validate our model.

We present a graph-based SLAM approach, using monocular vision and odometry, designed to operate on computationally constrained platforms. When computation and memory are limited, visual tracking becomes difficult or impossible, and map representation and update costs must remain low. Our system constructs a map of structured views using only weak temporal assumptions, and performs recognition and relative pose estimation over the set of views. Visual observations are fused with differential sensors in an incrementally optimized graph representation. Using variable elimination and constraint pruning, the graph complexity and storage is kept linear in explored space rather than in time. We evaluate performance on sequences with ground truth, and also compare to a standard graph SLAM approach.

Gathering a large collection of images has been made quite easy by social and image sharing websites, e.g. flickr.com. However, using such collections faces the problem that they contain a large number of duplicates and highly similar images. This work tackles the problem of how to automatically organize image collections into sets of similar images, called image families hereinafter. We thoroughly compare the performance of two approaches to measure image similarity: global descriptors vs. a set of local descriptors. We assess the performance of these approaches as the problem scales up to thousands of images and hundreds of families. We present our results on a new dataset of CD/DVD game covers.

This paper presents an extension to David Lowe’s well-known object recognition algorithm based on his Scale and Feature Invariant Transform (SIFT). One of the benefits of Lowe’s SIFT-based method is that it can recognize objects from only three keypoints. While this capability can be useful in circumstances where the cost of a false negative is high, it’s often the case that false positive are an equal, or greater, concern. We extend Lowe’s algorithm by adding the ability to use 3D constraints during matching. These constraints essentially eliminate false positive matches. We combine our extension with the original algorithm to retain the recognition power of the original method while adding significant robustness against false positives, thereby increasing overall classification power. In addition to improving recognition, our extension returns 3D pose information. Yet, it adds very little computational overhead to Lowe’s original algorithm. 1

This paper presents the Visual Simultaneous Localization and Mapping (vSLAMTM) algorithm, a novel algorithm for simultaneous localization and mapping (SLAM). The algorithm is vision-and odometry-based, and enables low-cost navigation in cluttered and populated environments. No initial map is required, and it satisfactorily handles dynamic changes in the environment, for example, lighting changes, moving objects and/or people. Typically, vSLAM recovers quickly from dramatic disturbances, such as “kidnapping”.

Conventional speech recognition engines extract Mel Frequency Cepstral Coefficients (MFCC) features from incoming speech. This paper presents a novel approach for feature extraction in which speech is processed according to the Auditory Image Model, a model of human psychoacoustics. We fist describe the proposed frontend, then we present recognition results obtained with the TIMIT database. Comparing with previously published results on the same task, the new approach achieves a 10% improvement in recognition accuracy.

In this paper we describe the need for, and the characteristics of a software architecture for commercial robotic products. We describe the Evolution Robotics Software Platform (ERSP/spl trade/), which provides a commercial-grade software architecture for mobile robots. The architecture has been designed to be modular, scalable, lightweight, portable, and reusable. It follows a hybrid model of data flow, combining behavior-based processing modules for real-time reactions with event-based task planning routines. We provide a detailed description of the main architectural components and some basic usage studies. We also highlight areas where compromises have been made and for which continuing work is needed.

This paper presents a system for automatic reconstruction of 3D structure using two unsynchronized cameras. Three images are acquired sequentially from the left, right, and again from the left camera. A virtual image from the left camera synchronized with the right image is created by interpolating matching points of interest (SIFT features) in the two left images. Both geometric and probabilistic criteria are used to select the correct set of matching features amongst the three views. In an indoor environment, the method typically results in 3D structure with approximately 200 feature points, with a median 3D accuracy of 1.6 cm when the average depth is 3 m and the robot has moved 1-2 cm between each image acquisition.

Service robotics products are becoming a reality. This paper describes three core technologies that enable the next generation of service robots. They are low-cost, make use of low-cost hardware, and prepare for a short time-to-market for product development. The first technology is an object recognition system, which can be used by the robot to interact with the environment The second technology is a vision-based navigation system (vSLAM/spl trade/), which simultaneously can build a map and localize the robot in the map. Finally, the third technology is a flexible and rich software platform (ERSP/spl trade/) that assists developers in rapid design and prototyping of robotics applications.

We propose a new camera-based biometric: visual signature identification. We discuss the importance of the parameterization of the signatures in order to achieve good classification results, independently of variations in the position of the camera with respect to the writing surface. We show that affine arc-length parameterization performs better than conventional time and Euclidean arc-length ones. We find that the system verification performance is better than 4 percent error on skilled forgeries and 1 percent error on random forgeries, and that its recognition performance is better than 1 percent error rate, comparable to the best camera-based biometrics.

The design and implementation of a camera-based, human-computer interface for acquisition of handwriting is presented. The camera focuses on a standard sheet of paper and images a common pen; the trajectory of the tip of the pen is tracked and the contact with the paper is detected. The recovered trajectory is shown to have sufficient spatio-temporal resolution and accuracy to enable handwritten character recognition. More than 100 subjects have used the system and have provided a large and heterogeneous set of examples showing that the system is both convenient and accurate.

Signatures can be acquired with a camera-based system with enough resolution to perform verification. This paper presents the performance of a visual-acquisition signature verification system, emphasizing on the importance of the parameterisation of the signature in order to achieve good classification results. A technique to overcome the lack of examples in order to estimate the generalization error of the algorithm is also described.

The problem of establishing correspondence and measuring the similarity of a pair of planar curves arises in many applications in computer vision and pattern recognition. This paper presents a new method for comparing planar curves and for performing matching at sub-sampling resolution. The analysis of the algorithm as well as its structural properties are described. The performance of the new technique applied to the problem of signature verification is shown and compared with the performance of the well-known Dynamic Time Warping algorithm.

Handwriting may be captured using a video camera, rather than the customary pressure-sensitive tablet. This paper presents a simple system based on correlation and recursive prediction methods that can track the tip of the pen in real time with sufficient spatio-temporal resolution and accuracy to enable handwritten character recognition. The system is tested on a large and heterogeneous set of examples and its performance is compared to that of three human operators and a commercial high-resolution pressure-sensitive tablet.

Handwriting may be captured using a video camera, rather than the customary pressure-sensitive tablet. This paper presents a simple system based on correlation and recursive prediction methods that can track the tip of the pen in real time with sufficient spatio-temporal resolution and accuracy to enable handwritten character recognition. The system is tested on a large and heterogeneous set of examples and its performance is compared to that of three human operators and a commercial high-resolution pressure-sensitive tablet.