Single Camera-Based Remote Physical Therapy: Verification on a Large Video Dataset

In recent years, several systems have been developed to capture human motion in real-time using common RGB cameras. This approach has great potential to become widespread among the general public as it allows the remote evaluation of exercise at no additional cost. The concept of using these systems in rehabilitation in the home environment has been discussed, but no work has addressed the practical problem of detecting basic body parts under different sensing conditions on a large scale. In this study, we evaluate the ability of the OpenPose pose estimation algorithm to perform keypoint detection of anatomical landmarks under different conditions. We infer the quality of detection based on the keypoint confidence values reported by the OpenPose. We used more than two thousand unique exercises for the evaluation. We focus on the influence of the camera view and the influence of the position of the trainees, which are essential in terms of the use for home exercise. Our results show that the position of the trainee has the greatest effect, in the following increasing order of suitability across all camera views: lying position, position on the knees, sitting position, and standing position. On the other hand, the effect of the camera view was only marginal, showing that the side view is having slightly worse results. The results might also indicate that the quality of detection of lower body joints is lower across all conditions than the quality of detection of upper body joints. In this practical overview, we present the possibilities and limitations of current camera-based systems in telerehabilitation.

Creation of the Simulator to Train Cosmonauts for Visual Instrumental Observations from the International Space Station Using Up-to-Date Information Technologies.

Visual Instrumental Observations (VIOs) of the Earth’s surface is a very special activity for cosmonauts that include searching, finding, and monitoring the objects with the unaided eye and their registration using optical facilities expanding capabilities of an operator’s visual analyzer. In order to impart the correct practical skills in performing VIOs from the ISS to cosmonauts in the course their ground training it is necessary to visualize the image observed by the naked eye and/or in the camera view finder taking into account an optical zoom, mutual location of a cosmonaut, a camera, a window, and the station at a certain point in orbit. For these purposes, the special simulator which simulates an external visual environment as close as possible to the flight conditions has been developed, that is the VIOs simulator designated to train cosmonauts for performing tasks in the field of geophysical studies and monitoring of the Earth.

Investigation of robust gait recognition for different appearances and camera view angles

A gait recognition framework is proposed to tackle the challenge of unknown camera view angles as well as appearance changes in gait recognition. In the framework, camera view angles are firstly identified before gait recognition. Two compact images, gait energy image (GEI) and gait modified Gaussian image (GMGI), are used as the base gait feature images. Histogram of oriented gradients (HOG) is applied to the base gait feature images to generate feature descriptors, and then a final feature map after principal component analysis (PCA) operations on the descriptors are used to train support vector machine (SVM) models for individuals. A set of experiments are conducted on CASIA gait database B to investigate how appearance changes and unknown view angles affect the gait recognition accuracy under the proposed framework. The experimental results have shown that the framework is robust in dealing with unknown camera view angles, as well as appearance changes in gait recognition. In the unknown view angle testing, the recognition accuracy matches that of identical view angle testing in gait recognition. The proposed framework is specifically applicable in personal identification by gait in a small company/organization, where unintrusive personal identification is needed.

NeLF: Neural Light-transport Field for Portrait View Synthesis and Relighting

Human portraits exhibit various appearances when observed from different views under different lighting conditions. We can easily imagine how the face will look like in another setup, but computer algorithms still fail on this problem given limited observations. To this end, we present a system for portrait view synthesis and relighting: given multiple portraits, we use a neural network to predict the light-transport field in 3D space, and from the predicted Neural Light-transport Field (NeLF)produce a portrait from a new camera view under a new environmental lighting. Our system is trained on a large number of synthetic models, and can generalize to different synthetic and real portraits under various lighting conditions. Our method achieves simultaneous view synthesis and relighting given multi-view portraits as the input, and achieves state-of-the-art results.

Mobile augmented reality based indoor map for improving geo-visualization

Unlike traditional visualization methods, augmented reality (AR) inserts virtual objects and information directly into digital representations of the real world, which makes these objects and data more easily understood and interactive. The integration of AR and GIS is a promising way to display spatial information in context. However, most existing AR-GIS applications only provide local spatial information in a fixed location, which is exposed to a set of problems, limited legibility, information clutter and the incomplete spatial relationships. In addition, the indoor space structure is complex and GPS is unavailable, so that indoor AR systems are further impeded by the limited capacity of these systems to detect and display location and semantic information. To address this problem, the localization technique for tracking the camera positions was fused by Bluetooth low energy (BLE) and pedestrian dead reckoning (PDR). The multi-sensor fusion-based algorithm employs a particle filter. Based on the direction and position of the phone, the spatial information is automatically registered onto a live camera view. The proposed algorithm extracts and matches a bounding box of the indoor map to a real world scene. Finally, the indoor map and semantic information were rendered into the real world, based on the real-time computed spatial relationship between the indoor map and live camera view. Experimental results demonstrate that the average positioning error of our approach is 1.47 m, and 80% of proposed method error is within approximately 1.8 m. The positioning result can effectively support that AR and indoor map fusion technique links rich indoor spatial information to real world scenes. The method is not only suitable for traditional tasks related to indoor navigation, but it is also promising method for crowdsourcing data collection and indoor map reconstruction.

ISDNet: AI-enabled Instance Segmentation of Aerial Scenes for Smart Cities

Aerial scenes captured by UAVs have immense potential in IoT applications related to urban surveillance, road and building segmentation, land cover classification, and so on, which are necessary for the evolution of smart cities. The advancements in deep learning have greatly enhanced visual understanding, but the domain of aerial vision remains largely unexplored. Aerial images pose many unique challenges for performing proper scene parsing such as high-resolution data, small-scaled objects, a large number of objects in the camera view, dense clustering of objects, background clutter, and so on, which greatly hinder the performance of the existing deep learning methods. In this work, we propose ISDNet (Instance Segmentation and Detection Network), a novel network to perform instance segmentation and object detection on visual data captured by UAVs. This work enables aerial image analytics for various needs in a smart city. In particular, we use dilated convolutions to generate improved spatial context, leading to better discrimination between foreground and background features. The proposed network efficiently reuses the segment-mask features by propagating them from early stages using residual connections. Furthermore, ISDNet makes use of effective anchors to accommodate varying object scales and sizes. The proposed method obtains state-of-the-art results in the aerial context.

Interactive Virtual Reality Touring System: A Case Study of Shulin Ji’an Temple in Taiwan

Most current tour guiding methods for Taiwanese temples employ graphic webpage frameworks combined with captioned pictures for introduction. This type of tour guiding lacks interactive presence. In addition, the audience may not be able to focus on browsing webpages or learn essential information from the introduction. This study adopted the Delphi method to evaluate the current developed system. This system was aimed at designing VR-based interaction that differs from conventional tour guiding methods to aid users in viewing the display space from their viewpoints. Users cannot only control camera view angles but also select the paths and guiding information as if they were walking in the temple. The analysis results revealed that, in general, the users perceived VR tour guiding as convenient and easy to use. The display and content of the tour guiding system presented clear information to the users, aiding them in gaining further understanding of the introduced item. Finally, the study results can serve as a reference for design research on VR applications in tour guiding.

Multi-Camera Vessel-Speed Enforcement by Enhancing Detection and Re-Identification Techniques

This paper presents a camera-based vessel-speed enforcement system based on two cameras. The proposed system detects and tracks vessels per camera view and employs a re-identification (re-ID) function for linking vessels between the two cameras based on multiple bounding-box images per vessel. Newly detected vessels in one camera (query) are compared to the gallery set of all vessels detected by the other camera. To train and evaluate the proposed detection and re-ID system, a new Vessel-reID dataset is introduced. This extensive dataset has captured a total of 2474 different vessels covered in multiple images, resulting in a total of 136,888 vessel bounding-box images. Multiple CNN detector architectures are evaluated in-depth. The SSD512 detector performs best with respect to its speed (85.0% Recall@95Precision at 20.1 frames per second). For the re-ID of vessels, a large portion of the total trajectory can be covered by the successful detections of the SSD model. The re-ID experiments start with a baseline single-image evaluation obtaining a score of 55.9% Rank-1 (49.7% mAP) for the existing TriNet network, while the available MGN model obtains 68.9% Rank-1 (62.6% mAP). The performance significantly increases with 5.6% Rank-1 (5.7% mAP) for MGN by applying matching with multiple images from a single vessel. When emphasizing more fine details by selecting only the largest bounding-box images, another 2.0% Rank-1 (1.4% mAP) is added. Application-specific optimizations such as travel-time selection and applying a cross-camera matching constraint further enhance the results, leading to a final 88.9% Rank-1 and 83.5% mAP performance.

A markerless augmented reality system for mobile devices

Augmented Reality (AR) combines a live camera view of a real world environment with computer-generated virtual content. Alignment of these viewpoints is done by recognizing artificial fiducial markers, or, more recently, natural features already present in the environment. This is known as Marker-based and Markerless AR respectively. We present a markerless AR system that is not limited to artificial markers, but is capable of rendering augmentations over user-selected textured surfaces, or ‘maps’. The system stores and differentiates between multiple maps, all created online. Once recognized, maps are tracked using a hybrid algorithm based on feature matching and inlier tracking. With the increasing ubiquity and capability of mobile devices, we believe it is possible to perform robust, markerless AR on current generation tablets and smartphones. The proposed system is shown to operate in real-time on mobile devices, and generate robust augmentations under a wide range of map compositions and viewing conditions.