Self-supervised monocular depth estimation (SSMDE) aims at predicting the dense depth maps of monocular images, by learning to minimize a photometric loss using spatially neighboring image pairs during training. While SSMDE offers a significant scalability advantage over supervised approaches, it performs poorly on reflective surfaces as the photometric constancy assumption of the photometric loss is violated. We note that the appearance of reflective surfaces is view-dependent and often there are views of such surfaces in the training data that are not contaminated by strong specular reflections. Thus, reflective surfaces can be accurately reconstructed by aggregating the predicted depth of these views. Motivated by this observation, we propose 3D distillation: a novel training framework that utilizes the projected depth of reconstructed reflective surfaces to generate reasonably accurate depth pseudo-labels. To identify those surfaces automatically, we employ an uncertainty-guided depth fusion method, combining the smoother and more accurate projected depth on reflective surfaces and the detailed predicted depth elsewhere. In our experiments using the ScanNet and 7-Scenes datasets, we show that 3D distillation not only significantly improves the prediction accuracy, especially on the problematic surfaces, but also that it generalizes well over various underlying network architectures and to new datasets.

Abstract Portrait synthesis creates realistic digital avatars which enable users to interact with others in a compelling way. Recent advances in StyleGAN and its extensions have shown promising results in synthesizing photorealistic and accurate reconstruction of human faces. However, previous methods often focus on frontal face synthesis and most methods are not able to handle large head rotations due to the training data distribution of StyleGAN. In this work, our goal is to take as input a monocular video of a face, and create an editable dynamic portrait able to handle extreme head poses. The user can create novel viewpoints, edit the appearance, and animate the face. Our method utilizes pivotal tuning inversion (PTI) to learn a personalized video prior from a monocular video sequence. Then we can input pose and expression coefficients to MLPs and manipulate the latent vectors to synthesize different viewpoints and expressions of the subject. We also propose novel loss functions to further disentangle pose and expression in the latent space. Our algorithm shows much better performance over previous approaches on monocular video datasets, and it is also capable of running in real‐time at 54 FPS on an RTX 3080.

Autonomous vehicles rely on precise high definition (HD) 3D maps for navigation. This paper presents the mapping component of an end-to-end system for crowdsourcing precise 3D maps with semantically meaningful landmarks such as traffic signs (6 dof pose, shape and size) and traffic lanes (3D splines). The system uses consumer grade parts, and in particular, relies on a single front facing camera and a consumer grade GPS. Using real-time sign and lane triangulation on-device in the vehicle, with offline sign/lane clustering across multiple journeys and offline Bundle Adjustment across multiple journeys in the backend, we construct maps with mean absolute accuracy at sign corners of less than 20 cm from 25 journeys. To the best of our knowledge, this is the first end-to-end HD mapping pipeline in global coordinates in the automotive context using cost effective sensors.

In this work, we present a new automatic system for scene reconstruction of high-level structural models. We start with identifying planar regions in depth images obtained with a SLAM system. Our main contribution is an approach which identifies constraints such as incidence and orthogonality of planar surfaces and uses them in an incremental optimization framework to extract high-level structural models. The result is a manifold mesh with a low number of polygons, immediately useful in many Augmented Reality applications such as inspection, interior design or spatial interaction.

We propose a novel solution to the generalized camera pose problem which includes the internal scale of the generalized camera as an unknown parameter. This further generalization of the well-known absolute camera pose problem has applications in multi-frame loop closure. While a well-calibrated camera rig has a fixed and known scale, camera trajectories produced by monocular motion estimation necessarily lack a scale estimate. Thus, when performing loop closure in monocular visual odometry, or registering separate structure-from-motion reconstructions, we must estimate a seven degree-of-freedom similarity transform from corresponding observations. Existing approaches solve this problem, in specialized configurations, by aligning 3D triangulated points or individual camera pose estimates. Our approach handles general configurations of rays and points and directly estimates the full similarity transformation from the 2D-3D correspondences. Four correspondences are needed in the minimal case, which has eight possible solutions. The minimal solver can be used in a hypothesize-and-test architecture for robust transformation estimation. Our solver also produces a least-squares estimate in the overdetermined case. The approach is evaluated experimentally on synthetic and real datasets, and is shown to produce higher accuracy solutions to multi-frame loop closure than existing approaches.

In this proposal we show case two 3D reconstruction systems running in real-time on a tablet equipped with a depth sensor. We believe that the proposed set of demonstrations will engage ISMAR attendees both in terms of tracking technology and user experience. Both demos show state-of-the art 3D reconstruction technology and give attendees a chance to try hands-on our tracking with simple and interactive user interfaces.

In this proposal we suggest a set of demonstrations to be presented at ISMAR 2014 around the topic of monocular tracking, mapping and rendering on mobile phones. We believe that the proposed set of demonstrations will engage ISMAR attendees both in terms of tracking technology and user experience. On the one hand, all demos show state-of-the art tracking and mapping technology and gives attendees a chance to try hands-on our tracking solutions and to discuss them with our researchers and developers. All demos also showcase compelling use cases enabled by our technology, encouraging discussions on topics such as integration of tracking technology with game engines and user experience research.

Augmented reality (AR) allows for an on-site presentation of information that is registered to the physical environment. Applications from civil engineering, which require users to process complex information, are among those which can benefit particularly highly from such a presentation. In this paper, we will describe how to use AR to support monitoring and documentation of construction site progress. For these tasks, the responsible staff usually requires fast and comprehensible access to progress information to enable comparison to the as-built status as well as to as-planned data. Instead of tediously searching and mapping related information to the actual construction site environment, our AR system allows for the access of information right where it is needed. This is achieved by superimposing progress as well as as-planned information onto the user's view of the physical environment. For this purpose, we present an approach that uses aerial 3-D reconstruction to automatically capture progress information and a mobile AR client for on-site visualization. Within this paper, we will describe in greater detail how to capture 3-D, how to register the AR system within the physical outdoor environment, how to visualize progress information in a comprehensible way in an AR overlay, and how to interact with this kind of information. By implementing such an AR system, we are able to provide an overview about the possibilities and future applications of AR in the construction industry.

We propose the combination of a keyframe-based monocular SLAM system and a global localization method. The SLAM system runs locally on a camera-equipped mobile client and provides continuous, relative 6DoF pose estimation as well as keyframe images with computed camera locations. As the local map expands, a server process localizes the keyframes with a pre-made, globally-registered map and returns the global registration correction to the mobile client. The localization result is updated each time a keyframe is added, and observations of global anchor points are added to the client-side bundle adjustment process to further refine the SLAM map registration and limit drift. The end result is a 6DoF tracking and mapping system which provides globally registered tracking in real-time on a mobile device, overcomes the difficulties of localization with a narrow field-of-view mobile phone camera, and is not limited to tracking only in areas covered by the offline reconstruction.

Micro aerial vehicles equipped with high-resolution cameras can be used to create aerial reconstructions of an area of interest. In that context automatic flight path planning and autonomous flying is often applied but so far cannot fully replace the human in the loop, supervising the flight on-site to assure that there are no collisions with obstacles. Unfortunately, this workflow yields several issues, such as the need to mentally transfer the aerial vehicle’s position between 2D map positions and the physical environment, and the complicated depth perception of objects flying in the distance. Augmented Reality can address these issues by bringing the flight planning process on-site and visualizing the spatial relationship between the planned or current positions of the vehicle and the physical environment. In this paper, we present Augmented Reality supported navigation and flight planning of micro aerial vehicles by augmenting the user’s view with relevant information for flight planning and live feedback for flight supervision. Furthermore, we introduce additional depth hints supporting the user in understanding the spatial relationship of virtual waypoints in the physical world and investigate the effect of these visualization techniques on the spatial understanding.

The combination of a cheap single-point laser range finder (LRF) device and cameras has become increasingly useful in recent research and industrial applications. While the single dot laser range finder can only provide depth for a single pixel in the observed image, it's cost and size can be useful for handheld devices or very lightweight robotic platforms. In this work, we propose two accurate calibration methods for determining the position and direction of the laser range finder with respect to the camera. Notably, we can determine the full calibration, even without observing the laser range finder observation point in the camera image. We evaluate both methods on synthetic and real data demonstrating their efficiency and good behavior under noise.

Verification of paper documents is an important part of checking a person's identity, authorization for access or simply establishing a trusted currency. Many documents such as passports or paper bills include holograms or other view-dependent elements that are difficult to forge and therefore are used to verify the genuineness of that document. View-dependent elements change their appearance based both on viewing direction and dominant light sources, thus it requires special knowledge and training to accurately distinguish original elements from forgeries. We present an interactive application for mobile devices that integrates the recognition of the documents with the interactive verification of view-dependent elements. The system recognizes and tracks the paper document, provides user guidance for view alignment and presents a stored image of the element's appearance depending on the current view of the document also recording user decisions. We describe how to model and capture the underlying spatially varying BRDF representation of view-dependent elements. Furthermore, we evaluate this approach within a user study and demonstrate that such a setup captures images that are recognizable and that can be correctly verified.

The development of new Simultaneous Localization and Mapping (SLAM) techniques is quickly advancing in research communities and rapidly transitioning into commercial products. Creating accurate and high-quality SLAM maps relies on a robust initialization process. However, the robustness and usability of SLAM initialization for end-users has often been disregarded. This paper presents and evaluates a novel tracking system for 6DOF pose tracking between a single keyframe and the current camera frame, without any prior scene knowledge. Our system is particularly suitable for SLAM initialization, since it allows 6DOF pose tracking in the intermediate frames before a wide-enough baseline between two keyframes has formed. We investigate how our tracking system can be used to interactively guide users in performing an optimal motion for SLAM initialization. However, our findings from a pilot study indicate that the need for such motion can be completely hidden from the user and outsourced to our tracking system. Results from a second user study show that letting our tracking system create a SLAM map as soon as possible is a viable and usable solution. Our work provides important insight for SLAM systems, showing how our novel tracking system can be integrated with a user interface to support fast, robust and user-friendly SLAM initialization.

In-situ 3D Modeling becomes increasingly prominent in current Augmented Reality research, particularly for mobile scenarios. However, real-time performance and qualitative modeling remain highly challenging. In this work, we propose a new interactive 3D modeling approach for indoor environments, combining an assistive user interface and constrained reconstruction with a device consisting of a single-point laser range finder and a camera. Using our system, a user pans around capturing a panorama of the environment, while simultaneously measuring the distance to a single point per frame. An automatic detection process estimates planes from these sparse 3D measurements. The user can highlight specific geometric features in the environment, such as 2- or 3-way corners, with simple gestures, adding more 3D points to the estimation. The segmented planes are refined using a constrained optimization, enforcing orthogonality and parallel constraints as well as minimizing the number of planes used in the reconstruction. Finally a volumetric space-carving approach determines the geometry of the environment. Our reconstruction approach can output highly accurate models built only from simple, clean geometry. To examine the quantitative performance of our approach, we run evaluations on both synthetic and real data.

Handling degenerate rotation-only camera motion is a challenge for keyframe-based simultaneous localization and mapping with six degrees of freedom. Existing systems usually filter corresponding keyframe candidates, resulting in mapping starvation and tracking failure. We propose to employ these otherwise discarded keyframes to build up local panorama maps registered in the 3D map. Thus, the system is able to maintain tracking during rotational camera motions. Additionally, we seek to actively associate panoramic and 3D map data for improved 3D mapping through the triangulation of more new 3D map features. We demonstrate the efficacy of our approach in several evaluations that show how the combined system handles rotation only camera motion while creating larger and denser maps compared to a standard SLAM system.

Augmented Reality systems that run interactively and in real time, using high quality graphical displays and sensational cues, can create the illusion of virtual objects appearing to be real. This paper presents the design, implementation, and evaluation of BurnAR, a novel demonstration which enables users to experience the illusion of seeing their own hands burning, which we achieve by overlaying virtual flames and smoke on their hands. Surprisingly, some users reported an involuntary warming sensation of their hands. Based on these comments, we hypothesized that stimulation of multiple sensory modalities presented in this AR environment can induce an involuntary experience in an additional sensory pathway: observation of virtual flames resulting in a heat sensation. This cross-modal transfer, known as virtual synesthe-sia, is a temporary experience which affects some people who are not synesthetes and only lasts for a short time during the illusory experience. To verify our hypothesis, we conducted an exploratory study where participants experienced the BurnAR demonstration under controlled conditions.

Virtual try-on applications have become popular because they allow users to watch themselves wearing different clothes without the effort of changing them physically. This helps users to make quick buying decisions and, thus, improves the sales efficiency of retailers. Previous solutions usually involve motion capture, 3D reconstruction or modeling, which are time consuming and not robust for all body poses. Our method avoids these steps by combining image-based renderings of the user and previously recorded garments. It transfers the appearance of a garment recorded from one user to another by matching input and recorded frames, image-based visual hull rendering, and online registration methods. Using images of real garments allows for a realistic rendering quality with high performance. It is suitable for a wide range of clothes and complex appearances, allows arbitrary viewing angles, and requires only little manual input. Our system is particularly useful for virtual try-on applications as well as interactive games.

Image-based visual hull rendering is a method for generating depth maps of a desired viewpoint from a set of silhouette images captured by calibrated cameras. It does not compute a view-independent data representation, such as a voxel grid or a mesh, which makes it particularly efficient for dynamic scenes. When users are captured, the scene is usually dynamic, but does not change rapidly because people move smoothly within a subsecond time frame. Exploiting this temporal coherence to avoid redundant calculations is challenging because of the lack of an explicit data representation. This paper analyzes the image-based visual hull algorithm to find intermediate information that stays valid over time and is, therefore, worth to make explicit. We then derive methods that exploit this information to improve the rendering performance. Our methods reduce the execution time by up to 25 percent. When the user's motions are very slow, reductions of up to 50 percent are achieved.

The quality and completeness of 3D models obtained by Structure-fromMotion (SfM) heavily depend on the image acquisition process. If the user gets feedback about the reconstruction quality already during the acquisition, he can optimize this process. The goal of this paper is to support a user during image acquisition by giving online feedback of the current reconstruction quality. We propose an online SfM method that integrates wide-baseline still-images in an online fashion into a consistent reconstruction and we derive a surface model given the SfM point cloud. To guide the user to scene parts that are captured not very well, we colour the mesh according to redundancy and resolution information. In the experiments, we show that our approach makes the final SfM result predictable already during image acquisition. The method is suited for large-scale reconstructions as obtained by flying micro aerial vehicles as well as on small indoor environments. We propose a method that supports a user in the acquisition process in two ways: (a) sparse online SfM with accuracy close to offline methods and (b) surface extraction and quality visualization. The workflow of our method is shown in Figure 1.

Augmented Reality (AR) applications on mobile devices like smartphones and tablet computers have become increasingly popular. In this paper, for the first time in the AR domain, we present: (1) the influence of different handheld displays and (2) the exocentric depth perception. Unlike egocentric depth perception, exocentric depth perception has not been investigated in AR.

We present a modified upright SURF feature descriptor for mobile phone GPUs. Our implementation called uSURF-ES is multiple times faster than a comparable CPU variant on the same device. Our results proof the feasibility of modern mobile graphics accelerators for GPGPU tasks especially for the detection phase in natural feature tracking used in Augmented Reality applications.

The focus of this workshop is on presenting, discussing and demonstrating recent tracking methods and applications that work well in practice and that show some superiority over state-of-the-art methods. Rather than focusing on pure novelty, this workshop encourages presentations that concentrate on complete systems and integrated approaches. The TMA workshop looks at pose tracking from an end-to-end point of view.

Augmented Reality systems that run interactively and in real time, using high quality graphical displays and sensational cues, can create the illusion of virtual objects appearing to be real. This paper presents the design and implementation of BurnAR, a demonstration which enables users to experience the illusion of seeing their own hands burning, which we achieve by overlaying virtual flames and smoke on their hands. Surprisingly, some users reported an involuntary warming sensation of their hands.

In this paper we present an approach for visualizing time-oriented data of dynamic scenes in an on-site AR view. Visualizations of time-oriented data have special challenges compared to the visualization of arbitrary virtual objects. Usually, the 4D data occludes a large part of the real scene. Additionally, the data sets from different points in time may occlude each other. Thus, it is important to design adequate visualization techniques that provide a comprehensible visualization. In this paper we introduce a visualization concept that uses overview and detail techniques to present 4D data in different detail levels. These levels provide at first an overview of the 4D scene, at second information about the 4D change of a single object and at third detailed information about object appearance and geometry for specific points in time. Combining the three levels of detail with interactive transitions such as magic lenses or distorted viewing techniques enables the user to understand the relationship between them. Finally we show how to apply this concept for construction site documentation and monitoring.

Magic lens and static peephole interfaces are used in numerous consumer mobile phone applications such as Augmented Reality browsers, games or digital map applications in a variety of contexts including public spaces. Interface performance has been evaluated for various interaction tasks involving spatial relationships in a scene. However, interface usage outside laboratory conditions has not been considered in depth in the evaluation of these interfaces.

We present an efficient natural feature tracking pipeline solely implemented in JavaScript. It is embedded in a web technology-based Augmented Reality system running plugin-free in web browsers. The evaluation shows that real-time framerates on desktop computers and while on smartphones interactive framerates are achieved.

Real-time three-dimensional acquisition of real-world scenes has many important applications in computer graphics, computer vision and human-computer interaction. Inexpensive depth sensors such as the Microsoft Kinect allow to leverage the development of such applications. However, this technology is still relatively recent, and no detailed studies on its scalability to dense and view-independent acquisition have been reported. This paper addresses the question of what can be done with a larger number of Kinects used simultaneously. We describe an interference-reducing physical setup, a calibration procedure and an extension to the KinectFusion algorithm, which allows to produce high quality volumetric reconstructions from multiple Kinects whilst overcoming systematic errors in the depth measurements. We also report on enhancing image based visual hull rendering by depth measurements, and compare the results to KinectFusion. Our system provides practical insight into achievable spatial and radial range and into bandwidth requirements for depth data acquisition. Finally, we present a number of practical applications of our system.

A convincing combination of virtual and real data in an Augmented Reality (AR) application requires detailed 3D information about the real world scene. In many situations extensive model data is not available, while sparse representations such as outlines on a map exist. In this paper, we present a novel approach using such sparse 3D model data to seed automatic image segmentation and infer a dense depth map of an environment. Sparse 3D models of known landmarks, such as points and lines from GIS databases, are projected into a registered image and initialize 2D image segmentation at the projected locations in the image. For the segmentation we propose different techniques, which combine shape information, semantics given by the database, and the visual appearance in the referenced image. The resulting depth information of objects in the scene can be used in many applications, including occlusion handling, label placement, and 3D modeling.

Summary form only given. We report on a disruption in organizational dynamics arising from the introduction of model-driven development tools in General Motors. The introduction altered the balance of collaboration deeply, and the organization is still negotiating with its aftermath. Our report illustrates one consequence of tool adoption in groups, and that these consequences should be understood to facilitate technical change.

Self-localization in large environments is a vital task for accurately registered information visualization in outdoor Augmented Reality (AR) applications. In this work, we present a system for self-localization on mobile phones using a GPS prior and an online-generated panoramic view of the user's environment. The approach is suitable for executing entirely on current generation mobile devices, such as smartphones. Parallel execution of online incremental panorama generation and accurate 6DOF pose estimation using 3D point reconstructions allows for real-time self-localization and registration in large-scale environments. The power of our approach is demonstrated in several experimental evaluations.

Virtual dressing rooms for the fashion industry and digital entertainment applications aim at creating an image or a video of a user in which he or she wears different garments than in the real world. Such images can be displayed, for example, in a magic mirror shopping application or in games and movies. Current solutions involve the error-prone task of body pose tracking. We suggest an approach that allows users who are captured by a set of cameras to be virtually dressed with previously recorded garments in 3D. By using image-based algorithms, we can bypass critical components of other systems, especially tracking based on skeleton models. We rather transfer the appearance of a garment from one user to another by image processing and image-based rendering. Using images of real garments allows for photo-realistic rendering quality with high performance.

Registration of real space and virtual information is a fundamental requirement for any augmented reality system. This paper presents an interactive method to quickly create a 3D room model and annotate locations within the room to provide registration anchors for virtual information. The method operates on a mobile phone and uses a visual rotation tracker to obtain orientation tracking for in-situ applications. The simple interaction allows non-expert users to create models of their environment and thus contribute marked-up representations to an online AR platform.

Rapid 3D reconstruction of environments has become an active research topic due to the importance of 3D models in a huge number of applications, be it in Augmented Reality (AR), architecture or other commercial areas. In this paper we present a novel system that allows for the generation of a coarse 3D model of the environment within several seconds on mobile smartphones. By using a very fast and flexible algorithm a set of panoramic images is captured to form the basis of wide field-of-view images required for reliable and robust reconstruction. A cheap on-line space carving approach based on Delaunay triangulation is employed to obtain dense, polygonal, textured representations. The use of an intuitive method to capture these images, as well as the efficiency of the reconstruction approach allows for an application on recent mobile phone hardware, giving visually pleasing results almost instantly.

We present a real-time camera pose tracking and mapping system which uses the assumption of a planar scene to implement a highly efficient mapping algorithm. Our light-weight mapping approach is based on keyframes and plane-induced homographies between them. We solve the planar reconstruction problem of estimating the keyframe poses with an efficient image rectification algorithm. Camera pose tracking uses continuously extended and refined planar point maps and delivers robustly estimated 6DOF poses. We compare system and method with bundle adjustment and monocular SLAM on synthetic and indoor image sequences. We demonstrate large savings in computational effort compared to the monocular SLAM system while the reduction in accuracy remains acceptable.

The S&T Program Chairs are delighted to welcome you to ISMAR 2011, the 10th symposium on Mixed and Augmented Reality! This year's symposium continues a long tradition of ISMAR meetings, a series that itself followed a related series of IWAR, ISMR, and ISAR meetings. This year's Science and Technology (S&T) track comprises a mix of highly selective mixed and augmented reality research and related work. Specifically, in the S&T program this year you will find 26 papers, 27 posters, as well as a stimulating mixture of keynote talks, demonstrations, tutorials, workshops and the tracking competition. All of these elements of the program are the result of dedicated hard work by members of various conference committees and additional volunteers, and we would like to thank all of them for their generous efforts.

Many mixed reality systems require the real-time capture and re-rendering of the real world to integrate real objects more closely with the virtual graphics. This includes novel view-point synthesis for virtual mirror or telepresence applications. For real-time performance, the latency between capturing the real world and producing the virtual output needs to be as little as possible. Image-based visual hull (IBVH) rendering is capable of rendering novel views from segmented images in real time. We improve upon existing IBVH implementations in terms of robustness and performance by reformulating the tasks of major components. Moreover, we enable high resolutions and little latency by exploiting view- and frame coherence. The suggested algorithm includes image warping between successive frames under the constraint of redraw volumes. These volumes form a boundary of the motion and deformation in the scene, and can be constructed efficiently by describing them as the visual hull of a set of bounding rectangles which are cast around silhouette differences in image-space. As a result, our method can handle arbitrarily moving and deforming foreground objects and free viewpoint motion at the same time, while still being able to reduce workload by reusing previous rendering results.

A common goal of outdoor augmented reality (AR) is the presentation of annotations that are registered to anchor points in the real world. We present an enhanced approach for registering and tracking such anchor points, which is suitable for current generation mobile phones and can also successfully deal with the wide variety of viewing conditions encountered in real life outdoor use. The approach is based on on-the-fly generation of panoramic images by sweeping the camera over the scene. The panoramas are then used for stable orientation tracking, while the user is performing only rotational movements. This basic approach is improved by several new techniques for the re-detection and tracking of anchor points. For the re-detection, specifically after temporal variations, we first compute a panoramic image with extended dynamic range, which can better represent varying illumination conditions. The panorama is then searched for known anchor points, while orientation tracking continues uninterrupted. We then use information from an internal orientation sensor to prime an active search scheme for the anchor points, which improves matching results. Finally, global consistency is enhanced by statistical estimation of a global rotation that minimizes the overall position error of anchor points when transforming them from the source panorama in which they were created, to the current view represented by a new panorama. Once the anchor points are redetected, we track the user's movement using a novel 3-degree-of-freedom orientation tracking approach that combines vision tracking with the absolute orientation from inertial and magnetic sensors. We tested our system using an AR campus guide as an example application and provide detailed results for our approach using an off-the-shelf smartphone. Results show that the re-detection rate is improved by a factor of 2 compared to previous work and reaches almost 90% for a wide variety of test cases while still keeping the ability to run at interactive frame rates.

Localization of mobile devices is an essential task in Augmented Reality and has therefore been an active research topic for many years. Typically, indoor tracking approaches, such as methods based on infrared or ultra-wide-band, require preparations of the environment and special hardware sensors. Conversely, image feature tracking approaches can provide orientation estimates without special tracking hardware installations. With the advent of mobile devices equipped with sensors such as digital cameras, image based localization gains importance in Augmented Reality. Typically, fiducial marker tracking was considered as a standard image based localization method. We propose the use of natural image feature based tracking methods, which are a generalization of the same principals but do not require the presence of fiducial tracking targets.

Magnetic compasses and accelerometers provide absolute orientation measurements within the earth's reference frame. However, sensor output typically suffers from jitter and external disturbances. Conversely, visual tracking provides more stable orientation estimation relative to an unknown initial orientation rather than to true north. We propose a 3-degree-of-freedom orientation tracking approach combining the accuracy and stability of vision tracking with the absolute orientation from inertial and magnetic sensors by estimating the offset between the initial orientation of the vision tracker and true north. We demonstrate that the approach improves absolute orientation estimation on a mobile phone device.

In augmented reality displays, X-Ray visualization techniques make hidden objects visible through combining the physical view with an artificial rendering of the hidden information. An important step in X-Ray visualization is to decide which parts of the physical scene should be kept and which should be replaced by overlays. The combination should provide users with essential perceptual cues to understand the relationship of depth between hidden information and the physical scene. In this paper we present an approach that addresses this decision in unknown environments by analyzing camera images of the physical scene and using the extracted information for occlusion management. Pixels are grouped into perceptually coherent image regions and a set of parameters is determined for each region. The parameters change the X-Ray visualization for either preserving existing structures or generating synthetic structures. Finally, users can customize the overall opacity of foreground regions to adapt the visualization.

In this paper, we present three techniques for 6DOF natural feature tracking in real time on mobile phones. We achieve interactive frame rates of up to 30 Hz for natural feature tracking from textured planar targets on current generation phones. We use an approach based on heavily modified state-of-the-art feature descriptors, namely SIFT and Ferns plus a template-matching-based tracker. While SIFT is known to be a strong, but computationally expensive feature descriptor, Ferns classification is fast, but requires large amounts of memory. This renders both original designs unsuitable for mobile phones. We give detailed descriptions on how we modified both approaches to make them suitable for mobile phones. The template-based tracker further increases the performance and robustness of the SIFT- and Ferns-based approaches. We present evaluations on robustness and performance and discuss their appropriateness for Augmented Reality applications.

Off-line model reconstruction relies on an image collection phase and a slow reconstruction phase, requiring a long time to verify a model obtained from an image sequence is acceptable. We propose a new model acquisition system, called ProFORMA, which generates a 3D model on-line as the input sequence is being collected. As the user rotates the object in front of a stationary camera, a partial model is reconstructed and displayed to the user to assist view planning. The model is also used by the system to robustly track the pose of the object. Models are rapidly produced through a Delaunay tetrahe-dralisation of points obtained from on-line structure from motion estimation, followed by a probabilistic tetrahedron carving step to obtain a textured surface mesh of the object. © 2009. The copyright of this document resides with its authors.

We present findings from field trials of MapLens, a mobile augmented reality (AR) map using a magic lens over a paper map. Twenty-six participants used MapLens to play a location-based game in a city centre. Comparisons to a group of 11 users with a standard 2D mobile map uncover phenomena that arise uniquely when interacting with AR features in the wild. The main finding is that AR features facilitate place-making by creating a constant need for referencing to the physical, and in that it allows for ease of bodily configurations for the group, encourages establishment of common ground, and thereby invites discussion, negotiation and public problem-solving. The main potential of AR maps lies in their use as a collaborative tool.

ProFORMA, an on-line reconstruction system for textured objects rotated by a user's hand, can be coupled with augmented reality (AR) to allow users to rapidly generate textured 3D models. We demonstrate how the use of an overlaid mesh model and 3D arrow can be used to assist the user in view planning, guiding the user to collect new keyframes from desirable views. The method described is particularly suited for use with AR headsets, providing guidance with minimal user input and allowing in situ modelling using the head-mounted camera (ProFORMA does not require a completely stationary camera).

Outdoor Augmented Reality typically requires tracking in unprepared environments. For global registration, Global Positioning System (GPS) is currently the best sensing technology, but its precision and update rate are not sufficient for high quality tracking. We present a system that uses Kalman filtering for fusion of Differential GPS (DGPS) or Real-Time Kinematic (RTK) based GPS with barometric heights and also for an inertial measurement unit with gyroscopes, magnetometers and accelerometers to improve the transient oscillation. Typically, inertial sensors are subjected to drift and magnetometer measurements are distorted by electro-magnetic fields in the environment. For compensation, we additionally apply a visual orientation tracker which is drift-free through online mapping of the unknown environment. This tracker allows for correction of distortions of the 3-axis magnetic compass, which increases the robustness and accuracy of the pose estimates. We present results of applying this approach in an industrial application scenario.

In this paper we present two techniques for natural feature tracking in real-time on mobile phones. We achieve interactive frame rates of up to 20 Hz for natural feature tracking from textured planar targets on current-generation phones. We use an approach based on heavily modified state-of-the-art feature descriptors, namely SIFT and Ferns. While SIFT is known to be a strong, but computationally expensive feature descriptor, Ferns classification is fast, but requires large amounts of memory. This renders both original designs unsuitable for mobile phones. We give detailed descriptions on how we modified both approaches to make them suitable for mobile phones. We present evaluations on robustness and performance on various devices and finally discuss their appropriateness for augmented reality applications.

Outdoor augmented reality systems often rely on GPS to cover large environments. Visual tracking approaches can provide more accurate location estimates but typically require a manual initialisation procedure. This paper describes the combination of both techniques to create an accurate localisation system that does not require any additional input for (re-)initialisation. The 2D GPS position together with average user height is used as an initial estimate for the visual tracking. The large gap in available GPS accuracy versus required accuracy for initialisation is overcome through a search procedure that tries to minimise search time by improving the likelihood of finding the correct estimate early. Re-initialisation of the visual tracking system after catastrophic failures is further improved by modelling the GPS error with a Gaussian process to provide a better estimate of the current location, thereby decreasing search time.

Unknown environments pose a particular challenge for augmented reality applications because the 3D models required for tracking, rendering and interaction are not available ahead of time. Consequently, authoring of AR content must take place on-line. This work describes a set of techniques to simplify the online authoring of annotations in unknown environments using a simultaneous localisation and mapping (SLAM) system. The point-based SLAM system is extended to specifically track and estimate high-level features indicated by the user. The automatic estimation of these complex landmarks by the system relieves the user from the burden of manually specifying the full 3D pose of annotations while improving accuracy. These properties are especially interesting for remote collaboration applications where either user interfaces on handhelds or camera control by the remote expert are limited.

This paper presents a model-based hybrid tracking system for outdoor augmented reality in urban environments enabling accurate, realtime overlays for a handheld device. The system combines several well-known approaches to provide a robust experience that surpasses each of the individual components alone: an edge-based tracker for accurate localisation, gyroscope measurements to deal with fast motions, measurements of gravity and magnetic field to avoid drift, and a back store of reference frames with online frame selection to re-initialize automatically after dynamic occlusions or failures. A novel edge-based tracker dispenses with the conventional edge model, and uses instead a coarse, but textured, 3D model. This yields several advantages: scale-based detail culling is automatic, appearance-based edge signatures can be used to improve matching and the models needed are more commonly available. The accuracy and robustness of the resulting system is demonstrated with comparisons to map-based ground truth data.

Paper-based cartographic maps provide highly detailed information visualisation with unrivalled fidelity and information density. Moreover, the physical properties of paper afford simple interactions for browsing a map or focusing on individual details, managing concurrent access for multiple users and general malleability. However, printed maps are static displays and while computer-based map displays can support dynamic information, they lack the nice properties of real maps identified above. We address these shortcomings by presenting a system to augment printed maps with digital graphical information and user interface components. These augmentations complement the properties of the printed information in that they are dynamic, permit layer selection and provide complex computer mediated interactions with geographically embedded information and user interface controls. Two methods are presented which exploit the benefits of using tangible artifacts for such interactions.

D applications, but with the focus lying on support for multi-processor and multi-pipeline systems, for distributed applications and for advanced rendering effects. Contrary to these developments, this work focusses on the expressiveness of the scene graph structure as a central tool for developing 3D user interfaces. We present the idea of a context for the traversal of a scene graph which allows to parameterize a scene graph and reuse it for different purposes. Such context sensitive scene graphs improve the inherent flexibility of a scene graph acting as a template with parameters bound during traversal. An implementation of this concept using an industry standard scene graph library is described and its use in a set of applications from the area of mobile augmented reality is demonstrated.

Current augmented reality (AR) applications require that the application software be written to support a specific AR interface set up. WireAR was developed to enable output from any OpenGL application to be viewed in an AR fashion. This enables the output from any legacy graphical or scientific visualization applications to be viewed in a collaborative AR setting. This demonstration shows how the output of standard desktop visualization programs can be embedded into an augmented reality experience.

This paper describes OpenTracker, an open software architecture that provides a generic solution to the different tasks involved in tracking input devices and processing tracking data for virtual environments. It combines a highly modular design with a configuration syntax based on XML, thus taking full advantage of this new technology. OpenTracker is a first attempt towards a "write once, track anywhere" approach to virtual reality application development.

This article describes OpenTracker, an open software architecture that provides a framework for the different tasks involved in tracking input devices and processing multi-modal input data in virtual environments and augmented reality application. The OpenTracker framework eases the development and maintenance of hardware setups in a more flexible manner than what is typically offered by virtual reality development packages. This goal is achieved by using an object-oriented design based on XML, taking full advantage of this new technology by allowing to use standard XML tools for development, configuration and documentation. The OpenTracker engine is based on a data flow concept for multi-modal events. A multi-threaded execution model takes care of tunable performance. Transparent network access allows easy development of decoupled simulation models. Finally, the application developer's interface features both a time-based and an event based model, that can be used simultaneously, to serve a large range of applications. OpenTracker is a first attempt towards a "'write once, input anywhere"' approach to virtual reality application development. To support these claims, integration into an existing augmented reality system is demonstrated. We also show how a prototype tracking equipment for mobile augmented reality can be assembled from consumer input devices with the aid of OpenTracker. Once development is sufficiently mature, it is planned to make Open-Tracker available to the public under an open source software license.