In this paper, we study the problem of shape part retrieval in the point cloud domain. Existing shape retrieval methods typically rely on the presence of a complete query object, but what if the part of interest is missing? We present the Part Retrieval Pipeline (PReP), which combines metric learning techniques with a trained classification model to evaluate the suitability of potential replacement parts from a database, within an application scenario aimed at circular economy. Through a progressively more difficult training procedure, PReP learns to recognize suitable parts based solely on shape context. Owing to its compact parameterization and low computational requirements, it can search a repository of tens of thousands of spare parts in just a few seconds. We also establish an alternative baseline approach for comparison, document the unique challenges associated with this task, and identify key design choices to address them.

We propose a novel point cloud U-Net diffusion architecture for 3-D generative modeling capable of generating high-quality and diverse 3-D shapes while maintaining fast generation times. Our network employs a dual-branch architecture, combining the high-resolution representations of points with the computational efficiency of sparse voxels. Our fastest variant outperforms all nondiffusion generative approaches on unconditional shape generation, the most popular benchmark for evaluating point cloud generative models, while our largest model achieves state-of-the-art results among diffusion methods, with a runtime approximately 70% of the previously state-of-the-art point-voxel diffusion (PVD), measured on the same hardware setting. Beyond unconditional generation, we perform extensive evaluations, including conditional generation on all categories of ShapeNet, demonstrating the scalability of our model to larger datasets, and implicit generation, which allows our network to produce high-quality point clouds on fewer timesteps, further decreasing the generation time. Finally, we evaluate the architecture's performance in point cloud completion and super-resolution. Our model excels in all tasks, establishing it as a state-of-the-art diffusion U-Net for point cloud generative modeling. The code is publicly available at https://github.com/JohnRomanelis/SPVD.

Solving jigsaw puzzles, a long-standing challenge in both human cognition and artificial intelligence, has seen significant progress with modern computer vision techniques. In this paper, we introduce a diffusion-based framework for jigsaw puzzle reconstruction, leveraging denoising diffusion models to iteratively refine piece placements. Unlike prior methods that rely on anchored reference pieces and relative positioning, our approach directly regresses absolute positions, making it more flexible and generalizable. Additionally, we extend puzzle-solving beyond square pieces by incorporating polygonal partitions and employ DDIM for efficient inference. Our modular pipeline is adaptable to various puzzle formulations, and we demonstrate its effectiveness by achieving state-of-the-art performance on the JPwLEG benchmark.

We propose a novel framework, Guided Sparse Point-Voxel Diffusion (G-SPVD), for Point Cloud generation guided from a single visual input - either an image or a rough hand-drawn sketch, both from an unknown viewing angle. G-SPVD combines a Vision Transformer with a Diffusion Model that iteratively forms a noisy set of points to match the requested input. Our quantitative evaluation demonstrates that our framework achieves state-of-the-art results compared to other methods in single-image reconstruction on the ShapeNet dataset. Moreover, despite the reduced information available in sketchbased inputs, our sketch-guided model still attains competitive reconstruction metrics. We present several qualitative results for both tasks to further illustrate the effectiveness of our method. Finally, we evaluate our method on unconditional generation, demonstrating that our model can generate shapes with quality and diversity on par with the current state-of-the-art. Our code will be released upon publication.

The problem of multi-view spatial calibration remains challenging, particularly in scenarios characterized by minimal overlapping regions across sensor views, a configuration for which robust solutions are not yet fully established. To systematically address this issue, we introduce SYNCH2-Cal, an evaluation framework designed to generate synthetic scenes containing fiducial and pattern-based calibration targets, alongside configurable camera viewpoints with precisely known spatial calibration parameters. Utilizing these synthetic datasets, we systematically evaluated several established methods for multi-view spatial calibration. Additionally, we extended the framework to incorporate the generation of 3D point clouds from the synthetic scenes, thereby enabling a rigorous assessment of various state-of-the-art point cloud registration algorithms, including traditional ICP-based and learned methods. The results obtained establish a valuable benchmark for the comparative evaluation of both multi-view calibration methods and 3D point cloud registration algorithms. Furthermore, they provide a foundation to guide future improvements of existing techniques as well as the development of next-generation algorithms.

Abstract Time‐varying meshes (TVMs), that is mesh sequences with varying connectivity, are a greatly versatile representation of shapes evolving in time, as they allow a surface topology to change or details to appear or disappear at any time during the sequence. This, however, comes at the cost of large storage size. Since 2003, there have been attempts to compress such data efficiently. While the problem may seem trivial at first sight, considering the strong temporal coherence of shapes represented by the individual frames, it turns out that the varying connectivity and the absence of implicit correspondence information that stems from it makes it rather difficult to exploit the redundancies present in the data. Therefore, efficient and general TVM compression is still considered an open problem. We describe and categorize existing approaches while pointing out the current challenges in the field and hint at some related techniques that might be helpful in addressing them. We also provide an overview of the reported performance of the discussed methods and a list of datasets that are publicly available for experiments. Finally, we also discuss potential future trends in the field.

Point Cloud Registration is a key step in tasks like Simultaneous Localization and Mapping and 3D Reconstruction. Methods that manage to use geometric information efficiently, show remarkable registration results in practice. Although these methods, in general, introduce intricate cost functions that prove challenging for general-purpose optimization algorithms. In this work, we introduce an efficient algorithm for solving the Generalized Least Squares problem, involved in the Generalized Iterative Closest Point, the most representative method of this class. Our method successfully combines the speed of the fastest algorithms with the accuracy of the most precise.

Autonomous Vehicles (AVs) have recently attracted considerable attention due to their potential to significantly reduce road accidents and improve people’s lives. However, they rely solely on the data collected by their mounted sensors to make predictions, which can lead to inaccurate results if a sensor becomes occluded or damaged. This issue can be addressed by employing Vehicle-to-Vehicle communication, which allows a Connected Autonomous Vehicle (CAV) to interact with other CAVs within its field of view and exchange information about their surrounding objects. Existing research on cooperative perception has primarily focused on clear weather scenarios, with limited exploration into adverse weather conditions. This paper demonstrates the necessity of Vehicle-to-Vehicle communication by showcasing its benefits in maintaining high accuracy under adverse weather conditions. A collaborative perception system is introduced and its performance in foggy weather scenarios is assessed to further improve adverse weather perception. The pipeline of the network combines state-of-the-art methods for accurate object detection. Specifically, with PointPillars as the backbone, the Spatial-wise Adaptive Feature Fusion method is used to aggregate information from different vehicles. The model is trained on the large-scale dataset OPV2V and evaluated on modified data to simulate fog. The experiments show that cooperative perception can maintain high detection accuracy even in challenging weather conditions. Finally, a comparative analysis of LiDAR detectors for cooperative perception in bad weather conditions is presented.

Gesture recognition is a tool to enable novel interactions with different techniques and applications, like Mixed Reality and Virtual Reality environments. With all the recent advancements in gesture recognition from skeletal data, it is still unclear how well state-of-the-art techniques perform in a scenario using precise motions with two hands. This paper presents the results of the SHREC 2024 contest organized to evaluate methods for their recognition of highly similar hand motions using the skeletal spatial coordinate data of both hands. The task is the recognition of 7 motion classes given their spatial coordinates in a frame-by-frame motion. The skeletal data has been captured using a Vicon system and pre-processed into a coordinate system using Blender and Vicon Shogun Post. We created a small, novel dataset with a high variety of durations in frames. This paper shows the results of the contest, showing the techniques created by the 5 research groups on this challenging task and comparing them to our baseline method.

Autonomous vehicles are expected to operate safely in real-life road conditions in the next years. Nevertheless, unanticipated events such as the existence of unexpected objects in the range of the road, can put safety at risk. The advancement of sensing and communication technologies and Internet of Things may facilitate the recognition of hazardous situations and information exchange in a cooperative driving scheme, providing new opportunities for the increase of collaborative situational awareness. Safe and unobtrusive visualization of the obtained information may nowadays be enabled through the adoption of novel Augmented Reality (AR) interfaces in the form of windshields. Motivated by these technological opportunities, we propose a saliency-based distributed, cooperative and rendering scheme for increasing the driver’s situational awareness through (i) automated negative obstacle (potholes) detection, (ii) AR visualization and (iii) information sharing (upcoming potential dangers) with other connected vehicles or road infrastructure. An extensive evaluation study using a variety of real datasets for pothole detection showed that the proposed method provides favorable results and features compared to other recent and relevant approaches.

The inference of 3D motion and dynamics of the human musculoskeletal system has traditionally been solved using physics-based methods that exploit physical parameters to provide realistic simulations. Yet, such methods suffer from computational complexity and reduced stability, hindering their use in computer graphics applications that require real-time performance. With the recent explosion of data capture (mocap, video) machine learning (ML) has started to become popular as it is able to create surrogate models harnessing the huge amount of data stemming from various sources, minimizing computational time (instead of resource usage), and most importantly, approximate real-time solutions. The main purpose of this paper is to provide a review and classification of the most recent works regarding motion prediction, motion synthesis as well as musculoskeletal dynamics estimation problems using ML techniques, in order to offer sufficient insight into the state-of-the-art and draw new research directions. While the study of motion may appear distinct to musculoskeletal dynamics, these application domains provide jointly the link for more natural computer graphics character animation, since ML-based musculoskeletal dynamics estimation enables modeling of more long-term, temporally evolving, ergonomic effects, while offering automated and fast solutions. Overall, our review offers an in-depth presentation and classification of ML applications in human motion analysis, unlike previous survey articles focusing on specific aspects of motion prediction.

Nowadays, three-dimensional (3D) meshes are widely used in various applications in different areas (e.g., industry, education, entertainment and safety). The 3D models are captured with multiple RGB-D sensors, and the sampled geometric manifolds are processed, compressed, simplified, stored, and transmitted to be reconstructed in a virtual space. These low-level processing applications require the accurate representation of the 3D models that can be achieved through saliency estimation mechanisms that identify specific areas of the 3D model representing surface patches of importance. Therefore, saliency maps guide the selection of feature locations facilitating the prioritization of 3D manifold segments and attributing to vertices more bits during compression or lower decimation probability during simplification, since compression and simplification are counterparts of the same process. In this work, we present a novel deep saliency mapping approach applied to 3D meshes, emphasizing decreasing the execution time of the saliency map estimation, especially when compared with the corresponding time by other relevant approaches. Our method utilizes baseline 3D importance maps to train convolutional neural networks. Furthermore, we present applications that utilize the extracted saliency, namely feature-aware multiscale compression and simplification frameworks.

With the advent of Connected and Autonomous Vehicles (CAVs) comes the very real risk that these vehicles will be exposed to cyber-attacks by exploiting various vulnerabilities. This paper gives a technical overview of the H2020 CARAMEL project (currently in the intermediate stage) in which Artificial Intelligent (AI)-based cybersecurity for CAVs is the main goal. Most of the possible scenarios are considered, by which an adversary can generate attacks on CAVs, such as attacks on camera sensors, GPS location, Vehicle to Everything (V2X) message transmission, the vehicle's On-Board Unit (OBU), etc. The counter-measures to these attacks and vulnerabilities are presented via the current results in the CARAMEL project achieved by implementing the designed security algorithms.

Hypertension is a chronic condition characterized by high pressure in the arteries of the human body. As a result, the heart is forced to work more intensively for the normal circulation of blood in the body. It is one of the most important risk factors for future fatal and non-cardiovascular diseases, stroke and kidney failure. In this article, Machine Learning (ML) is used to design effective models for predicting the long-term risk of older participants (over 50 years old) being diagnosed with hypertension. Our purpose is to train models with high sensitivity in identifying subjects at risk to avoid the future development and occurrence of hypertension following the proper interventions. In the context of the adopted methodology, two different class balancing methods are considered, under which features ranking is applied, and two ML models (namely, Decision tree and Naive Bayes) are compared based on Precision, Recall, F-Measure, Accuracy and Area Under Curve (AUC).

The new generation 3D scanner devices have revolutionized the way information from 3D objects is acquired, making the process of scene capturing and digitization straightforward. However, the effectiveness and robustness of conventional algorithms for real scene analysis are usually deteriorated due to challenging conditions, such as noise, low resolution, and bad perceptual quality. In this work, we present a methodology for identifying and registering partially-scanned and noisy 3D objects, lying in arbitrary positions in a 3D scene, with corresponding high-quality models. The methodology is assessed on point cloud scenes with multiple objects with large missing parts. The proposed approach does not require connectivity information and is thus generic and computationally efficient, thereby facilitating computationally demanding applications, like augmented reality. The main contributions of this work are the introduction of a layered joint registration and indexing scheme of cluttered partial point clouds using a novel multi-scale saliency extraction technique to identify distinctive regions, and an enhanced similarity criterion for object-to-model matching. The processing time of the process is also accelerated through 3D scene segmentation. Comparisons of the proposed methodology with other state-of-the-art approaches highlight its superiority under challenging conditions.

The acceptance of Automated Driving Systems is of key importance since it will determine whether they\n\nwill actually be used. This presentation describes contributions in this broad area from the perspective of the\n\nTrustonomy project with a focus on ethical decision support, human machine interfaces and trust assessment,\n\naimed at enhancing the experience of drivers and passengers in such vehicles.

Recent techniques of deep supervised learning have shown success in the domain of locomotion, providing high-quality and very realistic results. However, extending these systems to support further animation possibilities, with the character adapting to its environment and interacting with objects on a scene remains a challenging task. Manually specifying key frames to produce motion for individual interactions, can be a tedious and expensive task. In this paper, an inverse kinematic approach is discussed for collision avoidance with objects, as a way for extending the animation capabilities of characters. The system is based on the FABRIK algorithm in Unity3D for collision avoidance with objects of different sizes and geometry. The scalability of the system is demonstrated as it works with both quadruped and biped characters in real-time.

This paper presents a method that aims to achieve physics-based smoke rendering in VR despite its hardware limitations. This problem is addressed by decoupling the smoke animation and rendering into independent stages. Offline smoke diffusion simulation was conducted in respect to fluid dynamics. Real-Time rendering in VR was achieved by the combination of the simulation's accumulated data and light scattering theory. Results demonstrate that a satisfying framerate is accomplished making this approach suitable for developing immersive and interactive applications such as training simulations.

3D representations of highly deformable 3D models, such as dynamic 3D meshes, have recently become very popular due to their wide applicability in various domains. This trend inevitably leads to a demand for storage and transmission of voluminous data sets, making the need for the design of a robust and reliable compression scheme a necessity. In this work, we present an approach for dynamic 3D mesh compression, that effectively exploits the spatio-temporal coherence of animated sequences, achieving low compression ratios without noticeably affecting the visual quality of the animation. We show that, on contrary to mainstream approaches that either exploit spatial (e.g., spectral coding) or temporal redundancies (e.g., PCA-based method), the proposed scheme, achieves increased efficiency, by projecting the differential coordinates sequence to the subspace of the covariance of the point trajectories. An extensive evaluation study, using, different dynamic 3D models, highlights the benefits of the proposed approach in terms of both execution time and reconstruction quality, providing extremely low bit-per-vertex- per-frame (bpvf) rates.

Deep learning (DL) tends to be the integral part of Autonomous Vehicles (AVs). Therefore the development of scene analysis modules that are robust to various vulnerabilities such as adversarial inputs or cyber-attacks is becoming an imperative need for the future AV perception systems. In this paper, we deal with this issue by exploring the recent progress in Artificial Intelligence (AI) and Machine Learning (ML) to provide holistic situational awareness and eliminate the effect of the previous attacks on the scene analysis modules. We propose novel multi-modal approaches against which achieve robustness to adversarial attacks, by appropriately modifying the analysis Neural networks and by utilizing late fusion methods. More specifically, we propose a holistic approach by adding new layers to a 2D segmentation DL model enhancing its robustness to adversarial noise. Then, a novel late fusion technique has been applied, by extracting direct features from the 3D space and project them into the 2D segmented space for identifying inconsistencies. Extensive evaluation studies using the KITTI odometry dataset provide promising performance results under various types of noise.

The design of 3D object detection schemes that use point clouds as input in automotive applications has gained a lot of interest recently. Those schemes capitalize on Deep Neural Networks (DNNs) that have demonstrated impressive results in analyzing complex scenes. The proposed schemes are generally designed to improve the achieved performance, leading however to high performing approaches with high computational complexity. To mitigate this high complexity and to facilitate their deployment on edge devices, model compression and acceleration techniques can be utilized. In this paper, we propose compressed versions of two well-known 3D object detectors, namely, PointPillars and PV-RCNN, utilizing dictionary learning-based weight-sharing techniques. It is demonstrated that significant acceleration gains can be achieved with acceptable average precision loss when evaluated on the KITTI 3D object detection benchmark. These findings constitute a concrete step towards the deployment of high-performance networks in edge devices of limited resources, such as NVIDIA’s Jetson TX2.

Achieving interactivity for applications with physically-accurate caustics is proven to be a challenging task. We present a hybrid method utilizing ray tracing and rasterization which enables water caustic coverage to vast view distance in real-time. Inspired by photon mapping, we optimize the generation of photons using cascaded caustic maps to avoid tracing the first bounce of a ray from a light, replacing that step with rasterization. We introduce cascades as a set of caustic maps with varying resolution based on the distance from the viewer. In addition, since we adopt a splatting approach where each photon is rasterized into the image based on the extent of its contribution, we trace photon differentials in order to determine the size, shape and intensity of the splats so as to achieve adaptive anisotropic flux density estimation. Finally, to mitigate undersampled regions where lack of photons leads to noise, we propose the use of a cross-bilateral filter with an adaptive kernel radius. The radius is based on the perceived radiant energy of a photon relative to the scene’s luminance, and drastically improves performance. We demonstrate how the use of our method is able to perform interactive rendering of large-scale dynamic water caustics.

New generation 3-D scanning technologies are expected to create a revolution at the Industry 4.0, facilitating a large number of virtual manufacturing tools and systems. Such applications require the accurate representation of physical objects and/or systems achieved through saliency estimation mechanisms that identify certain areas of the 3-D model, leading to a meaningful and easier to analyze representation of a 3-D object. 3-D saliency mapping is, therefore, guiding the selection of feature locations and is adopted in a large number of low-level 3-D processing applications including denoising, compression, simplification, and registration. In this article, we propose a robust and fast method for creating 3-D saliency maps that accurately identifies sharp and small-scale geometric features in various industrial 3-D models. An extensive experimental study using a large number of 3-D scanned and CAD models verifies the effectiveness of the proposed method as compared to other recent and relevant approaches despite the constraints posed by complex geometry patterns or the presence of noise.

Recent advances in 3-D scanning technology have enabled the deployment of 3-D models in various industrial applications such as digital twins, remote inspection, and reverse engineering. Despite their evolving performance, 3-D scanners still introduce noise and artifacts in the acquired dense models. In this article, we propose a fast and robust denoising method for the dense 3-D scanned industrial models. The proposed approach employs conditional variational autoencoders to effectively filter face normals. Training and inference are performed in a sliding patch setup reducing the size of the required training data and execution times. We conducted extensive evaluation studies using 3-D scanned and CAD models. The results verify plausible denoising outcomes, demonstrating similar or higher reconstruction accuracy, compared to other state-of-the-art approaches. Specifically, for 3-D models with more than 1e4 faces, the presented pipeline is twice as fast as methods with equivalent reconstruction error.

Utilizing artificial intelligence (AI) for prediction of knee joint loading in lieu of computationally expensive musculoskeletal modeling is deemed necessary for real time assessment of pathological gait. In the current study, supervised learning was used to build a feed forward artificial neural network (FFANN) to learn the mapping from kinematics to joint force space. Joint angles and knee contact forces (KCFs) were estimated through inverse dynamics during treadmill walking at speeds ranging from 3 to 7 km/h. As a training set, 1848 trials of measured ground reaction forces along with calculated joint angles and KCFs were used from 40 participants. The network's predictive power was tested under different exposure levels to the full training set. Results indicate the increasing predictive power of the FFANN with increasing level of exposure to subject-specific data. Highest level of accuracy were found for medial KCF in the distal-proximal direction for all testing scenarios, with the highest Pearson coefficient R(mean across scenarios, R = 0.99) and lowest normalized root mean square error (mean across scenarios, NRMSE = 1.39%) when the model was trained using samples from all subjects. The predictive capability of the network was substantially reduced when all trials from the testing subject were excluded from the training set, although high predictive power (R = 0.87, NRMSE = 8.31%) for medial proximal-distal KCF was still observed. Findings suggest that surrogate biomechanical models produced by machine learning can predict KCFs with adequate accuracy levels, paving the way for real time analysis of gait and its consequent exploitation in clinical applications.

Sleep quality is an important factor highly impacting on quality of life, in particular in the case of older people who are still in the workforce and wish to remain professionally active, as it correlates to work efficiency and work ability. On one side, many self-reported sleep quality tools are used in the clinical practice, which are subject to the personal feelings of the individual about his/her sleep quality. On the other side, highly accurate and objective sleep assessment tools employed for diagnosis of sleep disorders are costly and can only be used for short periods of time in hospital setup. The aim of this paper is to present the SmartWork approach for continuous sleep quality assessment, which supports the triggering mechanisms for behavioral and lifestyle interventions in order to guide older people adopt healthier sleep habits and increase their sleep quality and satisfaction.

In view of the challenges of real-time 3D reconstruction and transmission, the research on tele-immersion systems has been quite intense. We present an end-to-end, realtime 3D reconstruction system of the human body's upper part in mixed reality applications, implemented with the use of a single depth camera on the capture side, whereas no special setup is required. Our system captures the scene, extracts the user's point cloud and by quantizing it, achieves real-time mesh generation and streaming. This system, together with the appropriate virtual (VR) or augmented (AR) reality equipment, creates a sense of a more direct, face-to-face communication, as if both users were in the same environment.

Throughout the years, several works have been proposed for 3D mesh denoising. Nevertheless, despite their reconstruction quality, there are still challenges related to the preservation of the fine surface features. Motivated by the impressive results of image denoising by 3D transform-domain collaborative filtering (CF), we extend it to 3D mesh denoising. CF is also capable of revealing the finest details shared by grouped blocks while preserving at the same time the unique features of each block. A new promising approach suggests unrolling the computational pipeline of CF into a convolutional neural network (CNN) structure increasing significantly the efficiency of this solution. In this paper, we successfully extend and apply this method to 3D meshes making a transition from face normals to pixels. Extensive evaluation studies carried out using a variety of 3D meshes verify that the proposed approach achieves plausible reconstruction outputs and provides very promising results.

Mesh saliency has been widely considered as the measure of visual importance of certain parts of 3D geometries, distinguishable from their surroundings, with respect to human visual perception. This work is based on the use of convolutional neural networks to extract saliency maps for large and dense 3D scanned models. The network is trained with saliency maps extracted by fusing local and global spectral characteristics. Extensive evaluation studies carried out using various 3D models, include visual perception evaluation in simplification and compression use cases. As a result, they verify the superiority of our approach as compared to other state-of-the-art approaches. Furthermore, these studies indicate that CNN-based saliency extraction method is much faster in large and dense geometries, allowing the application of saliency aware compression and simplification schemes in low-latency and energy-efficient systems.

The implications of frailty in older adults' health status and autonomy necessitates the understanding and effective management of this widespread condition as a priority for modern societies. Despite its importance, we still stand far from early detection, effective management and prevention of frailty. One of the most important reasons for this is the lack of sensitive instruments able to early identify frailty and pre-frailty conditions. The FrailSafe system provides a novel approach to this complex, medical, social and public health problem. It aspires to identify the most important components of frailty, construct cumulative metrics serving as biomarkers, and apply this knowledge and expertise for self-management and prevention. This paper presents a high-level overview of the FrailSafe system architecture providing details on the monitoring sensors and devices, the software front-ends for the interaction of the users with the system, as well as the back-end part including the data analysis and decision support modules. Data storage, remote processing and security issues are also discussed. The evaluation of the system by older individuals from 3 different countries highlighted the potential of frailty prediction strategies based on information and communication technology (ICT).

Obstructive inflammatory pulmonary diseases are life-long conditions of the airways affecting millions worldwide. A crucial step towards effective self-management is the adherence of the patients to their medication. Accurate detection of pressurised metered dose inhaler audio events can significantly improve medication adherence facilitating more meaningful interventions by medical personnel. Towards this direction, this work presents a data-driven approach for monitoring pressurised metered dose inhaler medication adherence employing recurrent neural networks with long short term memory (LSTM) units with spectrogram features. Evaluation studies took place with intra-subject and inter-subject settings, demonstrating that our LSTM-based approach yields higher accuracy compared to classical machine learning methods reaching a prediction accuracy ranging from 92% to 94% taking into account also samples comprised of a mixture of patterns belonging to more than one of the predefined classes.

Chronic inflammatory conditions are obstructive respiratory diseases that affect negatively the quality of life for patients and their families worldwide. The effective control of these diseases is achieved through the use of pressurized meter dose inhaler (pMDI). However, their management has been considered suboptimal, mainly due to the improper use of the inhaler device. Towards this direction, this work presents the use of deep sparse Convolutional Neural Network (CNN) as a classifier to provide a real-time assessment of medication adherence. The classification process is based only on recordings of inhaler use, achieving at the same time significantly lower computational complexity as compared to recent and relevant approaches, since no data transformation is needed. The proposed scheme reaches 95% classification accuracy, demonstrating that this method can also be executed on an embedded system dedicated to medication monitoring. Finally, timing studies were carried out and compared with other classification methods validating its computational efficiency.

Throughout the years, several works have been proposed for performing feature-preserving mesh denoising. Despite their reconstruction benefits, there are still challenges that need to be addressed. Meanwhile, several researchers, system engineers, and software developers have shown increasing interest in the application of deep learning approaches for performing several low-level information processing tasks, such as denoising, compression, etc., in image and video applications. In this paper, we present a method for reconstructing accurately static and dynamic noisy meshes by applying autoencoders on guided normals. To increase the effectiveness of the proposed method, we use different models for different set of faces that are classified as features and non-features. Extensive evaluation studies carried out using a variety of static and dynamic meshes, verify that the proposed approach achieves plausible reconstruction output despite the constraints posed by complex noise and geometry patterns.

Many mixed reality applications are based on the real-time compression and streaming of three-dimensional (3D) models. Thus, they demand very high-bandwidth and ultra-low latency from network specifications. The next-generation wireless networks will employ promising technologies to significantly improve the communication data rates. However, due to implementation complexity and thus increased energy consumption of these technologies, a trade-off between the quality-of-user-experience (QoE) and the hardware specifications is necessary. To overcome these limitations low-resolution quantizers have been of interest, which provide a trade-off between quality and complexity. In this paper, we propose a complexity-aware perceptual coding scheme that minimizes the reconstruction losses of the 3D models. Extensive simulations assuming different 3D models show that the proposed scheme achieves plausible reconstruction output offering significantly higher energy efficiency gains, as compared to a context unaware coding approaches.

The latest advancements in 3D scanning technologies have facilitated the generation and adoption of 3D models in several industrial applications ranging from manufacturing inspection and repair to digital twins and medical industry. All these applications require an accurate representation of physical objects through saliency mechanisms identifying certain areas of the 3D model that are considered as important information by humans. Hence, 3D saliency mapping is an essential mechanism in a number of 3D processing applications including denoising, compression, simplification, registration, viewpoint selection, etc. In this work, we propose a robust 3D saliency mapping method ideally suited for industrial 3D models with sharp and small scale geometric features. An extensive simulation study using a variety of 3D scanned and CAD models, verify the effectiveness of the proposed method as compared to other relevant approaches.

Through the years, several works have demonstrated high-quality 3D mesh denoising. Despite the high reconstruction quality, there are still challenges that need to be addressed ranging from variations in configuration parameters to high computational complexity. These drawbacks are crucial especially if the reconstructed models have to be used for quality check, inspection or repair in manufacturing environments where we have to deal with large objects resulting in very dense 3D meshes. Recently, deep learning techniques have shown that are able to automatically learn and find more accurate and reliable results, without the need for setting manually parameters. In this work, motivated by the aforementioned requirements, we propose a fast and reliable denoising method that can be effectively applied for reconstructing very dense noisy 3D models. The proposed method applies conditional variational autoencoders on face normals. Extensive evaluation studies carried out using a variety of 3D models verify that the proposed approach achieves plausible reconstruction outputs, very relative or even better of those proposed by the literature, in considerably faster execution times.

This paper presents multiple Mixed Reality 3D interaction, manipulation and simulation techniques in the context of the 2019 3DUI contest of the IEEE VR conference. The proposed schemes provide smart, seamless transition from the real to the virtual world and demonstrate passive haptics, mid-air haptics, object manipulation and abstract entities (time) manipulation. All techniques are integrated in the context of a mixed reality escape-room or treasure-hunt game, where information from both the real and the virtual world is necessary to solve the puzzle. The paper concludes with a discussion on the extensibility and translational application of the approaches in practical problem solving.

The increasing interest for reliable generation of large scale scenes and objects has facilitated several real-time applications. Although the resolution of the new generation geometry scanners are constantly improving, the output models, are inevitably noisy, requiring sophisticated approaches that remove noise while preserving sharp features. Moreover, we no longer deal exclusively with individual shapes, but with entire scenes resulting in a sequence of 3D surfaces that are affected by noise with different characteristics due to variable environmental factors (e.g., lighting conditions, orientation of the scanning device). In this work, we introduce a novel coarse-to-fine graph spectral processing approach that exploits the fact that the sharp features reside in a low dimensional structure hidden in the noisy 3D dataset. In the coarse step, the mesh is processed in parts, using a model based Bayesian learning method that identifies the noise level in each part and the subspace where the features lie. In the feature-aware fine step, we iteratively smooth face normals and vertices, while preserving geometric features. Extensive evaluation studies carried out under a broad set of complex noise patterns verify the superiority of our approach as compared to the state-of-the-art schemes, in terms of reconstruction quality and computational complexity.

Recently, there is a tremendous interest in the processing of unorganized point clouds, generated using a variety of 3D scanning technologies such as structured light and LIDAR systems. Without a doubt, the most compelling problem in this domain is the removal of outliers. To effectively address the aforementioned issue, we present a novel method, that detects accurately and efficiently the outliers by exploiting the spatial coherence in the object geometry and the sparsity of the outliers in the spatial domain. This is achieved by solving a convenient convex method called Robust PCA (RPCA). To demonstrate the effectiveness of the proposed technique, we evaluate it by using real scanned point clouds which are extremely dense consisting of millions of points.

Robotic vision and in particular 3D understanding has attracted intense research efforts the last few years due to its wide range of applications, such as robot-human interaction, augmented and virtual reality etc, and the introduction of lowcost 3D sensing devices. In this paper we explore one of the most popular problems encountered in 3D perception applications, namely the segmentation of a 3D scene and the retrieval of similar objects from a model database. We use a geometric approach for both the segmentation and the retrieval modules that enables us to develop a fast, low-memory footprint system without the use of large-scale annotated datasets. The system is based on the fast computation of surface normals and the encoding power of local geometric features. Our experiments demonstrate that such a complete 3D understanding framework is possible and advantages over other approaches as well as weaknesses are discussed.

There are several works performing accurate skeleton extraction, however, their main drawback is the extensive computational requirements and the lack of solutions that can be executed in multi-core computing systems. These challenges, become more demanding when we are dealing with dense 3D models. To cope with this scarcity we propose a novel method that extends a well known contraction-based skeletonization method, enabling its decentralization resulting in significant improvement in skeleton extraction times.

Recently, there has been increasing interest in the processing of dynamic scenes as captured by 3D scanners, ideally suited for challenging applications such as immersive tele-presence systems and gaming. Despite the fact that the resolution and accuracy of the modern 3D scanners are constantly improving, the captured 3D point clouds are usually noisy with a perceptive percentage of outliers, stressing the need of an approach with low computational requirements which will be able to automatically remove the outliers and create a consolidated point cloud. In this paper, we introduce a novel method which first recognizes and removes outliers from a dynamic point cloud sequence (DPCS) using a very fast Robust PCA (RPCA) approach and then we use a novel weighted Laplacian interpolation approach to achieve a fast and effective consolidation of a DPCS. Extensive evaluation studies, carried out using a collection of different DPCS, verify that the proposed technique achieves plausible reconstruction output despite the constraints posed by arbitrarily complex motion scenarios.

In this paper, we present a progressive compression scheme that enables aggressive compression ratios, by successfully identifying and encoding sharp and small scale geometric features. The accurate identification of the features is achieved by exploiting the low rank property of the captured geometry and the sparsity of the features in the Laplacian domain, permeating benefits from robust principal component analysis. Due to the visual importance of the identified geometric features, the geometry coding process, is optimized for preserving the geometric features at extremely low bit rates. Extensive evaluation studies, carried out using a collection of scanned and synthetic 3D models, show that the proposed feature aware high pass quantization method achieves extremely high compression ratios, offering at the same time meaningful approximations of the given surfaces. Finally, a short discussion regarding the applicability of the proposed feature identification scheme to smooth completion and feature preserving surface denoising is also offered.

Recently, there has been increasing interest for easy and reliable generation of 3-D animated models facilitating several real-time applications (like immersive telepresence, motion capture, and gaming). In most of these applications, the reconstruction of soft body animations is based on time-varying point clouds, which are nonuniformly sampled and highly incomplete. To overcome these significantly challenging imperfections without any additional information, first we introduce a novel reconstruction technique based on rank minimization theory, which can result in a unique solution to the otherwise ill-posed problem. This technique is further extended to exploit the spatial coherence, which usually characterizes the soft-body animations. Based on the developed tools, we propose a distributed consolidation technique where the reconstruction is performed by working simultaneously on several groups of frames. To achieve this, we impose temporal coherence between successive frame clusters by constraining the rank minimization problem. We validate the proposed techniques via experimental evaluation under different configurations and animated models, where we show that the high-frequency details of the models can be adequately recovered from a highly incomplete geometry data set.

Recently, there has been increasing interest for easy and reliable generation of 3D animated models facilitating several real-time applications. In most of these applications, the reconstruction of soft body animations is based on time-varying point clouds which are irregularly sampled and highly incomplete. To overcome these imperfections, we introduce a novel reconstruction technique, using graph-based matrix completion approaches. The presented method exploits spatio-temporal coherences by implicitly forcing the proximity of the adjacent 3D points in time and space. The proposed constraints are modeled by using the weighted Laplacian graphs and are constructed from the available points. Extensive evaluation studies, carried out using a collection of different highly-incomplete dynamic models, verify that the proposed technique achieves plausible reconstruction output despite the constraints posed by arbitrarily complex and motion scenarios.

With the growing demand for easy and reliable generation of 3D models representing real-world or synthetic objects, new schemes for acquisition, storage, and transmission of 3D meshes are required. In principle, 3D meshes consist of vertex positions and vertex connectivity. Vertex position encoders are much more resource demanding than connectivity encoders, stressing the need for novel geometry compression schemes. The design of an accurate and efficient geometry compression system can be achieved by increasing the compression ratio without affecting the visual quality of the object and minimizing the computational complexity. In this paper, we present novel compression/reconstruction schemes that enable aggressive compression ratios, without significantly reducing the visual quality. The encoding is performed by simply executing additions/subtractions. The benefits of the proposed method become more apparent as the density of the meshes increases, while it provides a flexible framework to trade efficiency for reconstruction quality. We derive a novel Bayesian learning algorithm that models the most significant graph Fourier transform coefficients of each submesh, as a multivariate Gaussian distribution. Then we evaluate iteratively the distribution parameters using the expectation-maximization approach. To improve the performance of the proposed approach in highly under determined problems, we exploit the local smoothness of the partitioned surfaces. Extensive evaluation studies, carried out using a large collection of different 3D models, show that the proposed schemes, as compared to the state-of-the-art approaches, achieve competitive compression ratios, offering at the same time significantly lower encoding complexity.

In this paper we propose an algorithm for multiple object tracking, a heavily researched but still challenging problem of computer vision. We follow the tracking by detection paradigm in an online fashion and formulate tracking as a typical assignment problem between detections and existing tracks that is solved by a modification of the Hungarian algorithm. Contrary to other methods that use a multitude of features based on appearance, optical flow and prior knowledge gained through training, we solely use clusters of point trajectories to link detections and tracks. Point trajectories are robust under partial occlusions and allow the expansion of a track even in the absence of a detection. At the core of our algorithm lies a motion segmentation method that extracts coherent clusters from triangulated point trajectories. Our algorithm achieves competitive results on the 2D MOT 2015 benchmark showcasing its potential.

Geometry contraction algorithms are introduced in this work to implement the diverse respiratory configurations of lung related diseases associated with airways obstructions. In addition, computational fluid dynamics (CFD) techniques along with fluid particle tracing (FPT) methods are utilized to efficiently evaluate the behavior of the airflow during the inhalation period, as well as to clarify the features of the inhaled particles in terms of regional deposition. Useful deductions are drawn regarding personalized medication in obstructed conditions.

Border gateway protocol (BGP) is the main protocol used on the Internet today, for the exchange of routing information between different networks. The lack of authentication mechanisms in BGP, render it vulnerable to prefix hijacking attacks, which raise serious security concerns regarding both service availability and data privacy. To address these issues, this study presents BGPGraph, a scheme for detecting and visualising Internet routing anomalies. In particular, BGPGraph introduces a novel BGP anomaly metric that quantifies the degree of anomaly on the BGP activity, and enables the analyst to obtain an overview of the BGP status. The analyst, is afterwards able to focus on significant time windows for further analysis, by using a hierarchical graph visualisation scheme. Furthermore, BGPGraph uses a novel method for the quantification of information visualisation that allows for the evaluation, and optimal selection of parameters, in case of the corresponding visual analytics algorithms. As a result, by utilising the proposed approach, four new BGP anomalies were able to be identified. Experimental demonstration in known BGP events, illustrates the significant analytics potential of the proposed approach in terms of identifying prefix hijacks and performing root cause analysis.

In order to produce realistic simulations and enhance immersion in augmented reality systems, solutions must not only present a realistic visual rendering of virtual objects, but also allow natural hand interactions. Most approaches capable of understanding user interaction with virtual content can often be restrictive or computationally expensive. To cope with these problems, we demonstrate a method which employs user's thumb and forefinger to interact with the virtual content in a natural way, utilizing a single RGB-D camera. Based on this method, we develop and realise an augmented reality chess game, focused on providing an immersive experience to users, so that they are able to manipulate virtual chess pieces seamlessly over a board of markers and play against a chess engine.

In this paper we present a methodology for creating a user-perspective view for mobile augmented reality (AR) systems. The most common video-see-through style for mobile AR systems is a device-perspective view, and the methods suggested for user-perspective rendering that are geometrically correct are very computationally expensive on current off-the-shelf devices. Based on this fact, we propose a real-time method which displays AR images according to the user's viewpoint by separating parts of the scene based on depth segmentation, and superimposing them as layers which gives the user a sense of motion parallax. With this paper, a prototype implementation on a desktop configuration is presented, accompanied by experimental results.

A novel big data building performance evaluation knowledge processing and mining system utilizing visual analytics is going to be presented in this paper. A large dataset comprised of building information, energy consumption, environmental measurements, human presence and behavior and business processes is going to be exploited for the building performance evaluation. Building performance evaluation is one of the most important factors in engineering that leads to building renovation and construction with low energy consumption and gas emissions in conjunction with comfort, utility and durability. For this purpose, business processes occurring in the building are correlated with the energy consumption and the human flows in the spatiotemporal domain modeling the dynamic behavior of the building. These models lead to the extraction of useful semantic information and the detection of spatiotemporal patterns that are important for the evaluation of the building performance. Furthermore, a number of novel visual analytics techniques allow the end-users to process data in different temporal resolutions and with different temporal filters, assisting them to detect patterns that may be difficult to be detected otherwise. The proposed visual analytics techniques support design and energy management decisions by visualizing the building measurements regarding business and comfort aspects. To do so, the proposed system includes a variety of techniques and components, properly selected to offer quick identification of focal points and evaluation of the building performance. Considering the increasing interest and the green building goals of almost all world governments including EU, the suggested methodology and application could be rendered a very useful tool for the Architecture and Engineering Community working on Building Performance Simulation and Analysis, and all related communities in Architect, Engineering and Construction (AEC) industry.

This paper presents a framework for the creation of virtual human behavioural profile which will account for physiological parameters of the body during motion activities. Initially, the subject's movements are captured along with the corresponding joint trajectories using a single Kinect sensor. To improve the fidelity of the motion trajectories a series of filters were implemented so as to diminish the effect of noise. Next, a musculoskeletal model of the human lower extremity with twenty degrees of freedom and 86 muscles is introduced and registered to the captured motion data through the inverse kinematics method. For the computational analysis and the estimation of joint torques during movements inverse and forward dynamics are utilized respectively. Finally, static optimization and computed muscle control are employed to estimate the muscle forces and muscle excitations necessary to perform the captured movements. Experimental results demonstrate the potential of the proposed approach, leading to numerous applications in the field of physics-based computer animation and virtual physiological human modeling.

This letter presents a novel probabilistic framework for augmenting the recognition performance of biometric systems with information from continuous soft biometric (SB) traits. In particular, by modelling the systematic error induced by the estimation of the SB traits, a modified efficient recognition probability can be extracted including information related both to the hard and SB traits. The proposed approach is applied without loss of generality in the case of gait recognition, where two state-of-the-art gait recognition systems are considered as hard biometrics and the height and stride length of the individuals are considered as SBs. Experimental validation on two known, large datasets illustrates significant advances in the recognition performance with respect to both identification and authentication rates.

This paper presents work conducted toward the biosignals-based automatic recognition of boredom, induced during video-game playing. For this purpose, common biosignal feature extraction methods were exploited and their capability to identify boredom was assessed. Moreover, for the first time, Legendre and Krawtchouk moments, as well as novel moment variations, were extracted as biosignal features and their potential toward automatic affect recognition was examined using the specific application scenario. The present analysis was conducted with ECG and GSR data collected from 19 different subjects, while boredom was naturally induced during the repetitive playing of a 3D video game. Conventional biosignal features as well as moment-based ones were found to be effective for the automatic recognition of boredom by achieving classification accuracies around 85 percent. Then, the joint use of moments and moment variations with conventional features was found to significantly improve classification accuracy by producing a maximum correct classification ratio of 94.17 percent.

This paper proposes a probabilistic framework for enhancing the recognition performance of 3D object queries based on soft shape descriptors, such as volume, surface and convexity. The goal is to improve the initial ranking results of pure geometry-based algorithms by capitalizing on the discriminative power of objects with close values on the so-called “soft shape descriptor”. The proposed method has been applied to benchmarking 3D object databases providing promising results demonstrating a clear improvement on the precision/recall ratio.

Abstract In this paper we present a new method for improving the performance of the widely used Bounding Volume Hierarchies for collision detection. The major contribution of our work is a culling algorithm that serves as a generalization of the Separating Axis Theorem for non parallel axes, based on the well‐known concept of support planes. We also provide a rigorous definition of support plane mappings and implementation details regarding the application of the proposed method to commonly used bounding volumes. The paper describes the theoretical foundation and an overall evaluation of the proposed algorithm. It demonstrates its high culling efficiency and in its application, significant improvement of timing performance with different types of bounding volumes and support plane mappings for rigid body simulations.

This letter presents a novel framework for gait recognition augmented with soft biometric information. Geometric gait analysis is based on Radon transforms and on gait energy images. User height and stride length information is extracted and utilized in a probabilistic framework for the detection of soft biometric features of substantial discrimination power. Experimental validation illustrates that the proposed approach for integrating soft biometric features in gait recognition advances significantly the identification and authentication performance.

In the above letter (ibid., vol. 17, no. 4, pp. 367-370, Apr. 10), equation (5) requires modification. The revised equation is presented here.

This paper presents a novel framework for partial matching and retrieval of 3-D models based on a query-by-range-image approach. Initially, salient features are extracted for both the query range image and the 3-D target model. The concept behind the proposed algorithm is that, for a 3-D object and a corresponding query range image, there should be a virtual camera with such intrinsic and extrinsic parameters that would generate an optimum range image, in terms of minimizing an error function that takes into account the salient features of the objects, when compared to other parameter sets or other target 3-D models. In the context of the developed framework, a novel method is also proposed to hierarchically search in the parameter space for the optimum solution. Experimental results illustrate the efficiency of the proposed approach even in the presence of noise or occlusion.

The present paper presents a novel approach for gait identification using 3D data and Krawtchouk moments to generate the descriptor feature vectors. The gait sequence is captured by a stereoscopic camera and the resulting 2.5D data are processed to generate a 3D hull of the captured silhouette. The 3D Protrusion Transform is then proposed that generates a silhouette image containing protrusion information. Finally, the descriptor vector of the extended silhouette is calculated using the Krawtchouk moments. Experimental evaluation illustrates that the proposed scheme is highly efficient in identifying gait sequences when compared to state of the art approaches.

This paper proposes an innovative gait identification and authentication method based on the use of novel 2-D and 3-D features. Depth-related data are assigned to the binary image silhouette sequences using two new transforms: the 3-D radial silhouette distribution transform and the 3-D geodesic silhouette distribution transform. Furthermore, the use of a genetic algorithm is presented for fusing information from different feature extractors. Specifically, three new feature extraction techniques are proposed: the two of them are based on the generalized radon transform, namely the radial integration transform and the circular integration transform, and the third is based on the weighted Krawtchouk moments. Extensive experiments carried out on USF ldquoGait Challengerdquo and proprietary HUMABIO gait database demonstrate the validity of the proposed scheme.

This paper presents a novel layered and fast framework for real-time collision detection and haptic interaction in virtual environments based on superquadric virtual object modeling. An efficient algorithm is initially proposed for decomposing the complex objects into subobjects suitable for superquadric modeling, based on visual salience and curvature constraints. The distance between the superquadrics and the mesh is then projected onto the superquadric surface, thus generating a distance map (SQ-Map). Approximate collision detection is then performed by computing the analytical equations and distance maps instead of triangle per triangle intersection tests. Collision response is then calculated directly from the superquadric models and realistic smooth force feedback is obtained using analytical formulae and local smoothing on the distance map. Experimental evaluation demonstrates that SQ-Map reduces significantly the computational cost when compared to accurate collision detection methods and does not require the huge amounts of memory demanded by distance field-based methods. Finally, force feedback is calculated directly from the distance map and the superquadric formulae.

The present paper describes an enhanced haptic virtual reality application for geometry education. The proposed application allows the user to create and edit a scene that consists of three-dimensional geometrical objects in order to form and solve complex geometrical problems. The core of the proposed scheme is based on a novel interference detection algorithm, which utilizes implicit surfaces, such as superquadrics, and their analytical description to speed up collision detection between the virtual hand and the virtual environment. Intersection tests are executed utilizing the implicit analytical formulae of the superquadrics. Experimental results demonstrate the high applicability of the proposed application and the huge gain in speed of the proposed collision detection approach when compared to state of the art methods.

This paper presents a novel object-based method for the generation of a stereoscopic image sequence from a monoscopic video, using bidirectional two-dimensional motion estimation for the recovery of rigid motion and structure and a Bayesian framework to handle occlusions. The latter is based on extended Kalman filters and an efficient method for reliably tracking object masks. Experimental results show that the layered object scene representation, combined with the proposed algorithm for reliably tracking object masks throughout the sequence, yields very accurate results.

This paper presents a novel pyramidal representation scheme for deformable object modelling, which uses a hierarchical approach to optimize the system's performance by executing the simulation in every level of a pyramid. The simulation results of each level are used to predict the lower level's state. The prediction is used as initial guess for the simulation of the lower level. The above procedure is repeated until the final level is reached. Experimental evaluation demonstrates that the proposed scheme is able to reduce significantly the computational cost, especially when the simulation involves procedures, which need large numerical computation like the conjugate gradient in implicit integration schemes.