Search

Wearable on-body millimeter-wave (mmWave) radars can provide obstacle detection and guidance for visually impaired people. However, their everyday performance is hindered by the rigid form factor and limited scanning range. In this article, we propose a low-profile, fast-scanning leaky-wave antenna (LWA) operating in the unlicensed V-band (57-64 GHz) to be integrated for on-body applications such as lightweight portable frequency-modulated continuous wave (FMCW) radars. The proposed LWA consists of meandering microstrips that can conform to the human body curvatures while maintaining beam-forming and beam-scanning properties. Experimental results demonstrate that the planar LWA achieves a realized gain above 10 dB with a fan-beam steering range in the H-plane from -40{\deg} to 43{\deg} over the operating frequency band while the half power beam-width (HPBW) is within 20{\deg}. Since for the foreseen application the antenna is supposed to conform to the user's body, the performance is also analyzed for a bent condition. The beam steering range changes to -32{\deg} to 50{\deg} when placed on the knee (corresponding to 80 mm radius). Under bending conditions, the LWA exhibits a maximum degradation of 1.75 dB, while the HPBW increases to 25{\deg}. This shows that due to the small size of the antenna, the impact of bending is low and the beam-forming and beam-scanning property of the designed LWA remain intact. Furthermore, we enable 2-D spatial scanning by employing an array of twelve LWAs with phased excitation, extending the scanning range in the E-plane from -40{\deg} to 40{\deg}, while the HPBW remains below 20{\deg} across the operational frequency range.

Cameras are essential vision instruments to capture images for pattern detection and measurement. Human-object interaction (HOI) detection is one of the most popular pattern detection approaches for captured human-centric visual scenes. Recently, Transformer-based models have become the dominant approach for HOI detection due to their advanced network architectures and thus promising results. However, most of them follow the one-stage design of vanilla Transformer, leaving rich geometric priors under-exploited and leading to compromised performance especially when occlusion occurs. Given that geometric features tend to outperform visual ones in occluded scenarios and offer information that complements visual cues, we propose a novel end-to-end Transformer-style HOI detection model, i.e., geometric features enhanced HOI detector (GeoHOI). One key part of the model is a new unified self-supervised keypoint learning method named UniPointNet that bridges the gap of consistent keypoint representation across diverse object categories, including humans. GeoHOI effectively upgrades a Transformer-based HOI detector benefiting from the keypoints similarities measuring the likelihood of human-object interactions as well as local keypoint patches to enhance interaction query representation, so as to boost HOI predictions. Extensive experiments show that the proposed method outperforms the state-of-the-art models on V-COCO and achieves competitive performance on HICO-DET. Case study results on the post-disaster rescue with vision-based instruments showcase the applicability of the proposed GeoHOI in real-world applications., Comment: Accepted to IEEE TIM

Close and continuous interaction with rich contacts is a crucial aspect of human activities (e.g. hugging, dancing) and of interest in many domains like activity recognition, motion prediction, character animation, etc. However, acquiring such skeletal motion is challenging. While direct motion capture is expensive and slow, motion editing/generation is also non-trivial, as complex contact patterns with topological and geometric constraints have to be retained. To this end, we propose a new deep learning method for two-body skeletal interaction motion augmentation, which can generate variations of contact-rich interactions with varying body sizes and proportions while retaining the key geometric/topological relations between two bodies. Our system can learn effectively from a relatively small amount of data and generalize to drastically different skeleton sizes. Through exhaustive evaluation and comparison, we show it can generate high-quality motions, has strong generalizability and outperforms traditional optimization-based methods and alternative deep learning solutions.

The radius valley, i.e., a dearth of planets with radii between 1.5 and 2 Earth radii, provides insights into planetary formation and evolution. Using homogenously revised planetary parameters from Kepler 1-minute short cadence light curves, we remodel transits of 72 small planets mostly orbiting low-mass stars, improving the precision and accuracy of planet parameters. By combining this sample with a similar sample of planets around higher-mass stars, we determine the depth of the radius valley as a function of stellar mass. We find that the radius valley is shallower for low-mass stars compared to their higher mass counterparts. Upon comparison, we find that theoretical models of photoevaporation under-predict the number of planets observed inside the radius valley for low-mass stars: with decreasing stellar mass, the predicted fraction of planets inside the valley remains approximately constant whereas the observed fraction increases. We argue that this provides evidence for the presence of icy planets around low-mass stars. Alternatively, planets orbiting low-mass stars undergo more frequent collisions and scatter in the stars' high-energy output may also cause planets to fill the valley. We predict that more precise mass measurements for planets orbiting low mass stars would be able to distinguish between these scenarios., Comment: 19 pages, accepted for publication in MNRAS

Purpose:Current methods for diagnosis of PD rely on clinical examination. The accuracy of diagnosis ranges between 73% and 84%, and is influenced by the experience of the clinical assessor. Hence, an automatic, effective and interpretable supporting system for PD symptom identification would support clinicians in making more robust PD diagnostic decisions. Methods: We propose to analyze Parkinson's tremor (PT) to support the analysis of PD, since PT is one of the most typical symptoms of PD with broad generalizability. To realize the idea, we present SPA-PTA, a deep learning-based PT classification and severity estimation system that takes consumer-grade videos of front-facing humans as input. The core of the system is a novel attention module with a lightweight pyramidal channel-squeezing-fusion architecture that effectively extracts relevant PT information and filters noise. It enhances modeling performance while improving system interpretability. Results:We validate our system via individual-based leave-one-out cross-validation on two tasks: the PT classification task and the tremor severity rating estimation task. Our system presents a 91.3% accuracy and 80.0% F1-score in classifying PT with non-PT class, while providing a 76.4% accuracy and 76.7% F1-score in more complex multiclass tremor rating classification task. Conclusion: Our system offers a cost-effective PT classification and tremor severity estimation results as warning signs of PD for undiagnosed patients with PT symptoms. In addition, it provides a potential solution for supporting PD diagnosis in regions with limited clinical resources.

Interaction-aware Autonomous Driving (IAAD) is a rapidly growing field of research that focuses on the development of autonomous vehicles (AVs) that are capable of interacting safely and efficiently with human road users. This is a challenging task, as it requires the autonomous vehicle to be able to understand and predict the behaviour of human road users. In this literature review, the current state of IAAD research is surveyed in this work. Commencing with an examination of terminology, attention is drawn to challenges and existing models employed for modelling the behaviour of drivers and pedestrians. Next, a comprehensive review is conducted on various techniques proposed for interaction modelling, encompassing cognitive methods, machine learning approaches, and game-theoretic methods. The conclusion is reached through a discussion of potential advantages and risks associated with IAAD, along with the illumination of pivotal research inquiries necessitating future exploration.

Here, using unique in-situ liquid secondary ion mass spectroscopy on isotope-labelled solid-electrolyte-interphase (SEI), assisted by cryogenic transmission electron microscopy and constrained ab initio molecular dynamics simulation, for the first time we answer the question regarding Li+ transport mechanism across SEI, and quantitatively determine the Li+-mobility therein. We unequivocally unveil that Li+ transport in SEI follows a mechanism of successive displacement, rather than "direct-hopping". We further reveal, in accordance with spatial-dependence of SEI structure across the thickness, the apparent Li+ self-diffusivity varies from 6.7*10-19 m2/s to 1.0*10-20 m2/s, setting a quantitative gauging of ionic transport behavior of SEI layer against the underlining electrode as well as the rate limiting step of battery operation. This direct study on Li+ kinetics in SEI fills part of the decade-long knowledge gap about the most important component in advanced batteries and provides more precise guidelines to the tailoring of interphasial chemistries for future battery chemistries., Comment: 25 pages, 4 figures

We present a software that predicts non-cleft facial images for patients with cleft lip, thereby facilitating the understanding, awareness and discussion of cleft lip surgeries. To protect patients privacy, we design a software framework using image inpainting, which does not require cleft lip images for training, thereby mitigating the risk of model leakage. We implement a novel multi-task architecture that predicts both the non-cleft facial image and facial landmarks, resulting in better performance as evaluated by surgeons. The software is implemented with PyTorch and is usable with consumer-level color images with a fast prediction speed, enabling effective deployment.

Learning view-invariant representation is a key to improving feature discrimination power for skeleton-based action recognition. Existing approaches cannot effectively remove the impact of viewpoint due to the implicit view-dependent representations. In this work, we propose a self-supervised framework called Focalized Contrastive View-invariant Learning (FoCoViL), which significantly suppresses the view-specific information on the representation space where the viewpoints are coarsely aligned. By maximizing mutual information with an effective contrastive loss between multi-view sample pairs, FoCoViL associates actions with common view-invariant properties and simultaneously separates the dissimilar ones. We further propose an adaptive focalization method based on pairwise similarity to enhance contrastive learning for a clearer cluster boundary in the learned space. Different from many existing self-supervised representation learning work that rely heavily on supervised classifiers, FoCoViL performs well on both unsupervised and supervised classifiers with superior recognition performance. Extensive experiments also show that the proposed contrastive-based focalization generates a more discriminative latent representation.

We investigate the effect of observing cadence on the precision of radius ratio values obtained from transit light curves by performing uniform Markov Chain Monte Carlo fits of 46 exoplanets observed by the Transiting Exoplanet Survey Satellite (TESS) in multiple cadences. We find median improvements of almost 50% when comparing fits to 20s and 120s cadence light curves to 1800s cadence light curves, and of 37% when comparing 600s cadence to 1800s cadence. Such improvements in radius precision are important, for example, to precisely constrain the properties of the radius valley or to characterize exoplanet atmospheres. We also implement a numerical Information Analysis to predict the precision of parameter estimates for different observing cadences. We tested this analysis on our sample and found it reliably predicts the effect of shortening observing cadence with errors in the predicted % precision of <0,5% for most cases. We apply this method to 157 TESS object of interest that have only been observed with 1800s cadence to predict the precision improvement that could be obtained by reobservations with shorter cadences and provide the full table of expected improvements. We report the 10 planet candidates that would benefit the most from reobservations at short cadence. Our implementation of the Information Analysis for the prediction of the precision of exoplanet parameters, Prediction of Exoplanet Precisions using Information in Transit Analysis (PEPITA) is made publicly available., Comment: 15 pages, 9 figures. Accepted for publication in MNRAS

The characteristics of the radius valley, i.e., an observed lack of planets between 1.5-2 Earth radii at periods shorter than about 100 days, provide insights into the formation and evolution of close-in planets. We present a novel view of the radius valley by refitting the transits of 431 planets using Kepler 1-minute short cadence observations, the vast majority of which have not been previously analysed in this way. In some cases, the updated planetary parameters differ significantly from previous studies, resulting in a deeper radius valley than previously observed. This suggests that planets are likely to have a more homogeneous core composition at formation. Furthermore, using support-vector machines, we find that the radius valley location strongly depends on orbital period and stellar mass and weakly depends on stellar age, with $\partial \log {\left(R_{p, \text{valley}} \right)}/ \partial \log{P} = -0.096_{-0.027}^{+0.023}$, $\partial \log {\left(R_{p, \text{valley}} \right)}/ \partial \log{M_{\star}} = 0.231_{-0.064}^{+0.053}$, and $\partial \log {\left(R_{p, \text{valley}} \right)}/ \partial \log{\left( \text{age} \right)} = 0.033_{-0.025}^{+0.017}$. These findings favour thermally-driven mass loss models such as photoevaporation and core-powered mass loss, with a slight preference for the latter scenario. Finally, this work highlights the value of transit observations with short photometric cadence to precisely determine planet radii, and we provide an updated list of precisely and homogeneously determined parameters for the planets in our sample., Comment: 19 pages, 17 figures, accepted in MNRAS

Early prediction is clinically considered one of the essential parts of cerebral palsy (CP) treatment. We propose to implement a low-cost and interpretable classification system for supporting CP prediction based on General Movement Assessment (GMA). We design a Pytorch-based attention-informed graph convolutional network to early identify infants at risk of CP from skeletal data extracted from RGB videos. We also design a frequency-binning module for learning the CP movements in the frequency domain while filtering noise. Our system only requires consumer-grade RGB videos for training to support interactive-time CP prediction by providing an interpretable CP classification result.

Musculoskeletal and neurological disorders are the most common causes of walking problems among older people, and they often lead to diminished quality of life. Analyzing walking motion data manually requires trained professionals and the evaluations may not always be objective. To facilitate early diagnosis, recent deep learning-based methods have shown promising results for automated analysis, which can discover patterns that have not been found in traditional machine learning methods. We observe that existing work mostly applies deep learning on individual joint features such as the time series of joint positions. Due to the challenge of discovering inter-joint features such as the distance between feet (i.e. the stride width) from generally smaller-scale medical datasets, these methods usually perform sub-optimally. As a result, we propose a solution that explicitly takes both individual joint features and inter-joint features as input, relieving the system from the need of discovering more complicated features from small data. Due to the distinctive nature of the two types of features, we introduce a two-stream framework, with one stream learning from the time series of joint position and the other from the time series of relative joint displacement. We further develop a mid-layer fusion module to combine the discovered patterns in these two streams for diagnosis, which results in a complementary representation of the data for better prediction performance. We validate our system with a benchmark dataset of 3D skeleton motion that involves 45 patients with musculoskeletal and neurological disorders, and achieve a prediction accuracy of 95.56%, outperforming state-of-the-art methods., Comment: Journal of Medical Systems

A Cleft lip is a congenital abnormality requiring surgical repair by a specialist. The surgeon must have extensive experience and theoretical knowledge to perform surgery, and Artificial Intelligence (AI) method has been proposed to guide surgeons in improving surgical outcomes. If AI can be used to predict what a repaired cleft lip would look like, surgeons could use it as an adjunct to adjust their surgical technique and improve results. To explore the feasibility of this idea while protecting patient privacy, we propose a deep learning-based image inpainting method that is capable of covering a cleft lip and generating a lip and nose without a cleft. Our experiments are conducted on two real-world cleft lip datasets and are assessed by expert cleft lip surgeons to demonstrate the feasibility of the proposed method., Comment: 4 pages, 2 figures, BHI 2022

Synthesizing multi-character interactions is a challenging task due to the complex and varied interactions between the characters. In particular, precise spatiotemporal alignment between characters is required in generating close interactions such as dancing and fighting. Existing work in generating multi-character interactions focuses on generating a single type of reactive motion for a given sequence which results in a lack of variety of the resultant motions. In this paper, we propose a novel way to create realistic human reactive motions which are not presented in the given dataset by mixing and matching different types of close interactions. We propose a Conditional Hierarchical Generative Adversarial Network with Multi-Hot Class Embedding to generate the Mix and Match reactive motions of the follower from a given motion sequence of the leader. Experiments are conducted on both noisy (depth-based) and high-quality (MoCap-based) interaction datasets. The quantitative and qualitative results show that our approach outperforms the state-of-the-art methods on the given datasets. We also provide an augmented dataset with realistic reactive motions to stimulate future research in this area. The code is available at https://github.com/Aman-Goel1/IMM, Comment: Accepted to SCA 2022 (will be published in CGF)

Parkinson's disease (PD) is a progressive neurodegenerative disorder that results in a variety of motor dysfunction symptoms, including tremors, bradykinesia, rigidity and postural instability. The diagnosis of PD mainly relies on clinical experience rather than a definite medical test, and the diagnostic accuracy is only about 73-84% since it is challenged by the subjective opinions or experiences of different medical experts. Therefore, an efficient and interpretable automatic PD diagnosis system is valuable for supporting clinicians with more robust diagnostic decision-making. To this end, we propose to classify Parkinson's tremor since it is one of the most predominant symptoms of PD with strong generalizability. Different from other computer-aided time and resource-consuming Parkinson's Tremor (PT) classification systems that rely on wearable sensors, we propose SPAPNet, which only requires consumer-grade non-intrusive video recording of camera-facing human movements as input to provide undiagnosed patients with low-cost PT classification results as a PD warning sign. For the first time, we propose to use a novel attention module with a lightweight pyramidal channel-squeezing-fusion architecture to extract relevant PT information and filter the noise efficiently. This design aids in improving both classification performance and system interpretability. Experimental results show that our system outperforms state-of-the-arts by achieving a balanced accuracy of 90.9% and an F1-score of 90.6% in classifying PT with the non-PT class., Comment: MICCAI 2022

Motion control algorithms in the presence of pedestrians are critical for the development of safe and reliable Autonomous Vehicles (AVs). Traditional motion control algorithms rely on manually designed decision-making policies which neglect the mutual interactions between AVs and pedestrians. On the other hand, recent advances in Deep Reinforcement Learning allow for the automatic learning of policies without manual designs. To tackle the problem of decision-making in the presence of pedestrians, the authors introduce a framework based on Social Value Orientation and Deep Reinforcement Learning (DRL) that is capable of generating decision-making policies with different driving styles. The policy is trained using state-of-the-art DRL algorithms in a simulated environment. A novel computationally-efficient pedestrian model that is suitable for DRL training is also introduced. We perform experiments to validate our framework and we conduct a comparative analysis of the policies obtained with two different model-free Deep Reinforcement Learning Algorithms. Simulations results show how the developed model exhibits natural driving behaviours, such as short-stopping, to facilitate the pedestrian's crossing.

Detecting human-object interactions is essential for comprehensive understanding of visual scenes. In particular, spatial connections between humans and objects are important cues for reasoning interactions. To this end, we propose a skeleton-aware graph convolutional network for human-object interaction detection, named SGCN4HOI. Our network exploits the spatial connections between human keypoints and object keypoints to capture their fine-grained structural interactions via graph convolutions. It fuses such geometric features with visual features and spatial configuration features obtained from human-object pairs. Furthermore, to better preserve the object structural information and facilitate human-object interaction detection, we propose a novel skeleton-based object keypoints representation. The performance of SGCN4HOI is evaluated in the public benchmark V-COCO dataset. Experimental results show that the proposed approach outperforms the state-of-the-art pose-based models and achieves competitive performance against other models., Comment: Accepted by IEEE SMC 2022

Nowadays, we are more and more reliant on Deep Learning (DL) models and thus it is essential to safeguard the security of these systems. This paper explores the security issues in Deep Learning and analyses, through the use of experiments, the way forward to build more resilient models. Experiments are conducted to identify the strengths and weaknesses of a new approach to improve the robustness of DL models against adversarial attacks. The results show improvements and new ideas that can be used as recommendations for researchers and practitioners to create increasingly better DL algorithms.

Identifying sleep stages and patterns is an essential part of diagnosing and treating sleep disorders. With the advancement of smart technologies, sensor data related to sleeping patterns can be captured easily. In this paper, we propose a Convolution Neural Network (CNN) architecture that improves the classification performance. In particular, we benchmark the classification performance from different methods, including traditional machine learning methods such as Logistic Regression (LR), Decision Trees (DT), k-Nearest Neighbour (k-NN), Naive Bayes (NB) and Support Vector Machine (SVM), on 3 publicly available sleep datasets. The accuracy, sensitivity, specificity, precision, recall, and F-score are reported and will serve as a baseline to simulate the research in this direction in the future.

Early diagnosis and intervention are clinically considered the paramount part of treating cerebral palsy (CP), so it is essential to design an efficient and interpretable automatic prediction system for CP. We highlight a significant difference between CP infants' frequency of human movement and that of the healthy group, which improves prediction performance. However, the existing deep learning-based methods did not use the frequency information of infants' movement for CP prediction. This paper proposes a frequency attention informed graph convolutional network and validates it on two consumer-grade RGB video datasets, namely MINI-RGBD and RVI-38 datasets. Our proposed frequency attention module aids in improving both classification performance and system interpretability. In addition, we design a frequency-binning method that retains the critical frequency of the human joint position data while filtering the noise. Our prediction performance achieves state-of-the-art research on both datasets. Our work demonstrates the effectiveness of frequency information in supporting the prediction of CP non-intrusively and provides a way for supporting the early diagnosis of CP in the resource-limited regions where the clinical resources are not abundant.

The wireless power transfer efficiency to implanted bioelectronic devices is constrained by several frequency-dependent physical mechanisms. Recent works have developed several mathematical formulations to understand these mechanisms and predict the optimal operating conditions. However, existing approaches rely on simplified body models, which are unable to capture important aspects of wireless power transfer. Therefore, this paper proposes the efficiency analysis approach in anatomical models that can provide insightful information on achieving the optimum operation conditions. First, this approach is validated with a theoretical spherical wave expansion analysis, and the results for a simplified spherical model and a human pectoral model are compared. The results show that although a magnetic receiver outperforms an electric one for near-field operation and both sources could be equally employed in far-field range, it is in mid-field that the maximum efficiency is achieved with an optimum frequency between 1-5 GHz depending on the implantation depth. The receiver orientation is another factor that affects the efficiency, with a maximum difference between the best and worst-case scenarios around five times for the electric source and over 13 times for the magnetic one. This approach is used to analyze the case of a deep-implanted pacemaker wirelessly powered by an on-body transmitter and subjected to stochastic misalignments. We evaluate the efficiency and exposure, and we demonstrate how a buffered transmitter can be tailored to achieve maximum powering efficiency. Finally, design guidelines that lead to optimal implantable wireless power transfer systems are established from the results obtained with the proposed approach.