In Situ Loading of Ni3ZnC0.7 Nanoparticles with Carbon Nanotubes to 3D Melamine Sponge Derived Hollow Carbon Skeleton toward Superior Microwave Absorption and Thermal Resistance.

The simple and low-cost construction of a 3D network structure is an ideal way to prepare high-performance electromagnetic wave (EMW) absorption materials. Herein, a series of carbon skeleton/carbon nanotubes/Ni3ZnC0.7 composites (CS/CNTs/Ni3ZnC0.7) are successfully prepared by in situ growth of Ni3ZnC0.7 and CNTs on 3D melamine sponge carbon. With the increase of precursor, Ni3ZnC0.7 nanoparticles nucleate and catalyze the generation of CNTs on the surface of the carbon skeleton. The minimum reflection loss (RL) value of the S60min composite (loading time of 60 min) reaches -86.6 dB at 1.6 mm and effective absorption bandwidth (EAB, RL≤-10 dB) is up to 9.3 GHz (8.7-18 GHz). The 3D network sponge carbon with layered micro/nanostructure and hollow skeleton promotes multiple reflection and absorption mechanisms of incident EMW. The N-doping and defects can be equivalent to an electric dipole, providing dipole polarization to increase dielectric relaxation. The uniform Ni3ZnC0.7 nanoparticles and CNTs play a key role in dissipating electromagnetic energy, blocking heat transfer, and enhancing the mechanical properties of the skeleton. Fortunately, the composite displays a quite low thermal conductivity of 0.09075 W m·K-1 and good flexibility, which can provide insulation and quickly recover to its original state after being stressed.

LightPRA: A Lightweight Temporal Convolutional Network for Automatic Physical Rehabilitation Exercise Assessment.

Research evidence shows that physical rehabilitation exercises prescribed by medical experts can assist in restoring physical function, improving life quality, and promoting independence for physically disabled individuals. In response to the absence of immediate expert feedback on performed actions, developing a Human Action Evaluation (HAE) system emerges as a valuable automated solution, addressing the need for accurate assessment of exercises and guidance during physical rehabilitation. Previous HAE systems developed for the rehabilitation exercises have focused on developing models that utilize skeleton data as input to compute a quality score for each action performed by the patient. However, existing studies have focused on improving scoring performance while often overlooking computational efficiency. In this research, we propose LightPRA (Light Physical Rehabilitation Assessment) system, an innovative architectural solution based on a Temporal Convolutional Network (TCN), which harnesses the capabilities of dilated causal Convolutional Neural Networks (CNNs). This approach efficiently captures complex temporal features and characteristics of the skeleton data with lower computational complexity, making it suitable for real-time feedback provided on resource-constrained devices such as Internet of Things (IoT) devices and Edge computing frameworks. Through empirical analysis performed on the University of Idaho-Physical Rehabilitation Movement Data (UI-PRMD) and KInematic assessment of MOvement for remote monitoring of physical REhabilitation (KIMORE) datasets, our proposed LightPRA model demonstrates superior performance over several state-of-the-art approaches such as Spatial-Temporal Graph Convolutional Network (STGCN) and Long Short-Term Memory (LSTM)-based models in scoring human activity performance, while exhibiting lower computational cost and complexity.

Genome-wide association study of growth curve parameters reveals novel genomic regions and candidate genes associated with metatarsal bone traits in chickens.

The growth and development of chicken bones have an enormous impact on the health and production performance of chickens. However, the development pattern and genetic regulation of the chicken skeleton are poorly understood. This study aimed to evaluate metatarsal bone growth and development patterns in chickens via non-linear models, and to identify the genetic determinants of metatarsal bone traits using a genome-wide association study (GWAS) based on growth curve parameters. Data on metatarsal length (MeL) and metatarsal circumference (MeC) were obtained from 471 F2 chickens (generated by crossing broiler sires, derived from a line selected for high abdominal fat, with Baier layer dams) at 4, 6, 8, 10, and 12 weeks of age. Four non-linear models (Gompertz, Logistic, von Bertalanffy, and Brody) were used to fit the MeL and MeC growth curves. Subsequently, the estimated growth curve parameters of the mature MeL or MeC (A), time-scale parameter (b), and maturity rate (K) from the non-linear models were utilized as substitutes for the original bone data in GWAS. The Logistic and Brody models displayed the best goodness-of-fit for MeL and MeC, respectively. Single-trait and multi-trait GWASs based on the growth curve parameters of the Logistic and Brody models revealed 4 618 significant single nucleotide polymorphisms (SNPs), annotated to 332 genes, associated with metatarsal bone traits. The majority of these significant SNPs were located on Gallus gallus chromosome (GGA) 1 (167.433-176.318 Mb), GGA2 (96.791-103.543 Mb), GGA4 (65.003-83.104 Mb) and GGA6 (64.685-95.285 Mb). Notably, we identified 12 novel GWAS loci associated with chicken metatarsal bone traits, encompassing 35 candidate genes. In summary, the combination of single-trait and multi-trait GWASs based on growth curve parameters uncovered numerous genomic regions and candidate genes associated with chicken bone traits. The findings benefit an in-depth understanding of the genetic architecture underlying metatarsal growth and development in chickens.

Conserved patterns and locomotor-related evolutionary constraints in the hominoid vertebral column.

The evolution of the hominoid lineage is characterized by pervasive homoplasy, notably in regions such as the vertebral column, which plays a central role in body support and locomotion. Few isolated and fewer associated vertebrae are known for most fossil hominoid taxa, but identified specimens indicate potentially high levels of convergence in terms of both form and number. Homoplasy thus complicates attempts to identify the anatomy of the last common ancestor of hominins and other taxa and stymies reconstructions of evolutionary scenarios. One way to clarify the role of homoplasy is by investigating constraints via phenotypic integration, which assesses covariation among traits, shapes evolutionary pathways, and itself evolves in response to selection. We assessed phenotypic integration and evolvability across the subaxial (cervical, thoracic, lumbar, sacral) vertebral column of macaques (n = 96), gibbons (n = 77), chimpanzees (n = 92), and modern humans (n = 151). We found a mid-cervical cluster that may have shifted cranially in hominoids, a persistent thoracic cluster that is most marked in chimpanzees, and an expanded lumbosacral cluster in hominoids that is most expanded in gibbons. Our results highlight the highly conserved nature of the vertebral column. Taxa appear to exploit existing patterns of integration and ontogenetic processes to shift, expand, or reduce cluster boundaries. Gibbons appear to be the most highly derived taxon in our sample, possibly in response to their highly specialized locomotion.

Structural Support for Large to Total Nasal Reconstruction.

When large defects of the nose are present, it is imperative to address all 3 layers: the external skin envelope, the osteocartilaginous support, and the inner mucosal lining. The middle structural framework is the primary factor in determining the overall shape of the nose, in addition to facilitating a functional and patent airway. As such, its reconstruction must be robust enough to provide lasting osteocartilaginous support while minimizing disfiguring bulk. The goal is replacement of missing tissue with grafts of similar strength, size, and shape. This article will review approaches to the reconstruction of structural support in large nasal defects.

Lathyrane and premyrsinane Euphorbia diterpenes against Alzheimer's disease: Bioinspired synthesis, anti-cholinesterase and neuroprotection bioactivity.

The first systematic acylated diversification of naturally scarce premyrsinane diterpenes, together with their biosynthetic precursors lathyrane diterpene were carried out. Two new series of premyrsinane derivates (1a-32a) and lathyrane derivates (1-32) were synthesized from the naturally abundant lathyrane diterpene Euphorbia factor L3 through a bioinspired approach. The cholinesterase inhibitory and neuroprotective activities of these diterpenes were investigated to explore potential anti-Alzheimer's disease (AD) bioactive lead compounds. In general, the lathyrane diterpenes showed the better acetylcholinesterase (AChE) inhibitory activity than that of premyrsinanes. The lathyrane derivative 17 bearing a 3-dimethylaminobenzoyl moiety showed the best AChE inhibition effect with the IC50 value of 7.1 µM. Molecular docking demonstrated that 17 could bond with AChE well (-8 kal/mol). On the other hand, premyrsinanes showed a better neuroprotection profile against H2O2-induced injury in SH-SY5Y cells. Among them, the premyrsinane diterpene 16a had significant neuroprotective effect with the cell viability rate of 113.5 % at 12.5 µM (the model group with 51.2 %). The immunofluorescence, western blot and reactive oxygen species (ROS) analysis were conducted to demonstrate the mechanism of 16a. Furthermore, a preliminary SAR analysis of the two categories of diterpenes was performed to provide the insights for anti-AD drug development.

Sc3+ substituted Na3V2(PO4)3 on N-doped porous carbon skeleton boosting high structural stability and superior electrochemical performance for full sodium ion batteries.

The poor structural stability and conductivity of Na3V2(PO4)3 (NVP) have been serious limitations to its development. In this paper, Sc3+ is selected to replace partial site of V3+ which can enhance its ability to bond with oxygen, forming the ScO6 octahedral unit, resulting in improved structural stability and better kinetic properties for the NVP system. Moreover, due to the larger ionic radius of Sc3+ compared to V3+, moderate Sc3+ substitution can support the crystal framework as pillar ions and expand the migration channels for de-intercalation of Na+, thus efficiently promoting ionic conductivity. The introduction of polyacrylonitrile (PAN) to provide an N-doped porous carbon substrate is another key aspect. The low-cost carbon resource of PAN can induce a beneficial nitrogen-doped carbon skeleton with defects, enhancing electronic conductivity at the interface to reduce the polarization phenomenon. The established pore structure can serve as a buffer for unit cell deformation caused by Na+ migration. Furthermore, the enlarged specific surface area provides more active sites for electrolyte infiltration, improving the material utilization rate. The after cycling X-ray Diffraction/scanning electron microscope (XRD/SEM) further confirms the stabilized porous carbon skeleton and improved crystal stability of Sc-3 material. Ex-situ XRD analysis shows that the crystal volume change in the Sc-3 cathode is relatively slight but reversible during the charge/discharge process, indicating that Sc3+ doping plays a crucial role in stabilizing the unit cell structure. The hybrid Sc/VO6 and PO4 units jointly build a strong bone structure to resist stress and weaken deformation. Accordingly, the optimized Sc-3 sample reveals an initial capacity of 115.9 mAh/g at 0.1C, with a capacity retention of 78.6 % after 2000 cycles at 30C. The Sc-3//CHC full battery can release a capacity of 191.3 mAh/g at 0.05C, accompanied by successful illumination, showcasing its promising practical applications.

Automated Pavement Condition Index Assessment with Deep Learning and Image Analysis: An End-to-End Approach.

The degradation of road pavements due to environmental factors is a pressing issue in infrastructure maintenance, necessitating precise identification of pavement distresses. The pavement condition index (PCI) serves as a critical metric for evaluating pavement conditions, essential for effective budget allocation and performance tracking. Traditional manual PCI assessment methods are limited by labor intensity, subjectivity, and susceptibility to human error. Addressing these challenges, this paper presents a novel, end-to-end automated method for PCI calculation, integrating deep learning and image processing technologies. The first stage employs a deep learning algorithm for accurate detection of pavement cracks, followed by the application of a segmentation-based skeleton algorithm in image processing to estimate crack width precisely. This integrated approach enhances the assessment process, providing a more comprehensive evaluation of pavement integrity. The validation results demonstrate a 95% accuracy in crack detection and 90% accuracy in crack width estimation. Leveraging these results, the automated PCI rating is achieved, aligned with standards, showcasing significant improvements in the efficiency and reliability of PCI evaluations. This method offers advancements in pavement maintenance strategies and potential applications in broader road infrastructure management.

Ionic Covalent Organic Frameworks in Adsorption and Catalysis.

The ion extraction and electro/photo catalysis are promising methods to address environmental and energy issues. Covalent organic frameworks (COFs) are a class of promising template to construct absorbents and catalysts because of their stable frameworks, high surface areas, controllable pore environments, and well-defined catalytic sites. Among them, ionic COFs as unique class of crystalline porous materials, with charges in the frameworks or along the pore walls, have shown different properties and resulting performance in these applications with those from charge-neutral COFs. In this review, current research progress based on the ionic COFs for ion extraction and energy conversion, including cationic/anionic materials and electro/photo catalysis is reviewed in terms of the synthesis strategy, modification methods, mechanisms of adsorption and catalysis, as well as applications. Finally, we demonstrated the current challenges and future development of ionic COFs in design strategies and applications.

Novel hybrid rigid-deformable fetal modeling for simulating the vaginal delivery within the second stage of labor.

BACKGROUND AND OBJECTIVE: The fetal representation as a 3D articulated body plays an essential role to describe a realistic vaginal delivery simulation. However, the current computational solutions have been oversimplified. The objective of the present work was to develop and evaluate a novel hybrid rigid-deformable modeling approach for the fetal body and then simulate its interaction with surrounding fetal soft tissues and with other maternal pelvis soft tissues during the second stage of labor. METHODS: CT scan data was used for 3D fetal skeleton reconstruction. Then, a novel hybrid rigid-deformable model of the fetal body was developed. This model was integrated into a maternal 3D pelvis model to simulate the vaginal delivery. Soft tissue deformation was simulated using our novel HyperMSM formulation. Magnetic resonance imaging during the second stage of labor was used to impose the trajectory of the fetus during the delivery. RESULTS: Our hybrid rigid-deformable fetal model showed a potential capacity for simulating the movements of the fetus along with the deformation of the fetal soft tissues during the vaginal delivery. The deformation energy density observed in the simulation for the fetal head fell within the strain range of 3 % to 5 %, which is in good agreement with the literature data. CONCLUSIONS: This study developed, for the first time, a hybrid rigid-deformation modeling of the fetal body and then performed a vaginal delivery simulation using MRI-driven kinematic data. This opens new avenues for describing more realistic behavior of the fetal body kinematics and deformation during the second stage of labor. As perspectives, the integration of the full skeleton body, especially the upper and lower limbs will be investigated. Then, the completed model will be integrated into our developed next-generation childbirth training simulator for vaginal delivery simulation and associated complication scenarios.

A rapid and simple procedure for the whole-mount bone staining of small fish.

The cartilage and bone structure has provided insightful knowledge about evolution and ecology of fish, which is an important component of biological oceanography. However, the whole-body bone staining is a lengthy and complicated process that typically takes five days to several months, and the improvement of the conventional method has been one of the important issues in this field. Here we report a quick and easy whole-mount bone staining method for small fish, in which a newly designed fixative is applied. Compared to conventional methods, this novel protocol is a straightforward process that could be adopted for small estuarine fish and other small vertebrates.

Nasal septum-derived chondroprogenitor cells control mandibular condylar resorption consequent to orthognathic surgery: a clinical trial.

Condylar resorption is an aggressive and disability form of temporomandibular joint (TMJ) degenerative disease, usually non-respondent to conservative or minimally invasive therapies and often leading to surgical intervention and prostheses implantation. This condition is also one of the most dreaded postoperative complications of orthognathic surgery, with severe cartilage erosion and loss of subchondral bone volume and mineral density, associated with a painful or not inflammatory processes. Because regenerative medicine has emerged as an alternative for orthopedic cases with advanced degenerative joint disease, we conducted a phase I/IIa clinical trial (U1111-1194-6997) to evaluate the safety and efficacy of autologous nasal septal chondroprogenitor cells. Ten participants underwent biopsy of the nasal septum cartilage during their orthognathic surgery. The harvested cells were cultured in vitro and analyzed for viability, presence of phenotype markers for mesenchymal stem and/or chondroprogenitor cells, and the potential to differentiate into chondrocytes, adipocytes, and osteoblasts. After the intra-articular injection of the cell therapy, clinical follow-up was performed using the Diagnostic Criteria for Temporomandibular Disorders (DC/TMD) and computed tomography (CT) images. No serious adverse events related to the cell therapy injection were observed during the 12-month follow-up period. It was found that autologous chondroprogenitors reduced arthralgia, promoted stabilization of mandibular function and condylar volume, and regeneration of condylar tissues. This study demonstrates that chondroprogenitor cells from the nasal septum may be a promise strategy for the treatment of temporomandibular degenerative joint disease that do not respond to other conservative therapies.

Traumatic fractures in an early 19th century museum skeleton suggest the homicide of an old Munich character: the history of "Finessensepperl" (Finesse Joseph).

The well preserved skeleton of Joseph Huber, a very well-known historical character of the 19th century Munich, also nicknamed "Finessen-Sepperl", is the starting point of the reconstruction of life and death of this historical individual. He was known as a postilion d´amour (love's messenger) of the Royal Bavarian capital with numerous comments and anecdotes and a few biographical sketches that indicate he remained well until the last few years of his life where requests for his duties lessened. The skeleton shows a small-sized male individual with almost complete loss of teeth, but otherwise very well-mineralized bone, having suffered from three episodes of trauma - an old-healed incomplete femoral neck fracture leading to severe osteoarthrosis, a clavicle fracture of the medial third with a few weeks old callus formation, and fresh serial rib fractures along with severe skull trauma with fractures of the os temporale and petrosum, presumably leading to intracranial bleeding and finally death. The type and distribution of these latter two injuries are in agreement with a murderous attack - which was retrospectively reported several years after his death, while the old-healed femoral neck fracture may have caused reduction in Joseph´s walking activities but not reduced requests for his services. Paleopathology not only identifies the terminal decline, but also previous diseases of this Old Bavarian character and thereby completes his story.

Ultrasound supports clinical decision-making in determining the Sanders' skeletal maturity score of the hand.

PURPOSE: The Sanders Scoring System has revolutionized the way we assess the remaining growth potential of the skeleton. However, because it involves radiation exposure, it must be used with caution in children. The purpose of the study was to evaluate whether the Sanders skeletal maturity score (SMS) could be accurately determined using ultrasound (U). METHODS: We took radiographs (R) of the hand and performed U of the thumb and index finger in 115 patients between six and 19 years of age who were undergoing treatment for scoliosis or limb deformities. Paediatric orthopaedic surgeons, a paediatrician, and a paediatric radiologist were evaluated the blinded images. Those classified images are based on the SMS and the Thumb Ossification Composite Index (TOCI). RESULTS: Intrarater reliability was high for SMS and slightly weaker for TOCI, but still significant. Interrater reliability was clear for R and weaker for U in both staging systems. Ultimately, SMS 3 and 7 achieved the highest percentage of concordance (P) of 71.7% and 66.0%, respectively, when U was performed. Combining the clinically relevant groups of SMS 3&4 and SMS 7&8 also significantly increased peak scores (SMS 3 and 4 P = 76.7%; SMS 7 and 8 P = 79.7%). The probabilities of peak scores were significantly weaker when the TOCI score was examined. CONCLUSION: Our study shows that U can be used effectively especially to measure stages 3 and 4 and stages 7 and 8 of SMS. The U method is easy to use and therefore may offer advantages in clinical practice without the need for radiation exposure.

A novel approach to the detection of facial wrinkles: Database, detection algorithm, and evaluation metrics.

Skin wrinkles result from intrinsic aging processes and extrinsic influences, including prolonged exposure to ultraviolet radiation and tobacco smoking. Hence, the identification of wrinkles holds significant importance in skin aging and medical aesthetic investigation. Nevertheless, current methods lack the comprehensiveness to identify facial wrinkles, particularly those that may appear insignificant. Furthermore, the current assessment techniques neglect to consider the blurred boundary of wrinkles and cannot differentiate images with varying resolutions. This research introduces a novel wrinkle detection algorithm and a distance-based loss function to identify full-face wrinkles. Furthermore, we develop a wrinkle detection evaluation metric that assesses outcomes based on curve, location, and gradient similarity. We collected and annotated a dataset for wrinkle detection consisting of 1021 images of Chinese faces. The dataset will be made publicly available to further promote wrinkle detection research. The research demonstrates a substantial enhancement in detecting subtle wrinkles through implementing the proposed method. Furthermore, the suggested evaluation procedure effectively considers the indistinct boundaries of wrinkles and is applicable to images with various resolutions.

Baicalin administration could rescue high glucose-induced craniofacial skeleton malformation by regulating neural crest development.

Hyperglycemia in pregnancy can increase the risk of congenital disorders, but little is known about craniofacial skeleton malformation and its corresponding medication. Our study first used meta-analysis to review the previous findings. Second, baicalin, an antioxidant, was chosen to counteract high glucose-induced craniofacial skeleton malformation. Its effectiveness was then tested by exposing chicken embryos to a combination of high glucose (HG, 50 mM) and 6 µM baicalin. Third, whole-mount immunofluorescence staining and in situ hybridization revealed that baicalin administration could reverse HG-inhibited neural crest cells (NCC) delamination and migration through upregulating the expression of Pax7 and Foxd3, and mitigate the disordered epithelial-mesenchymal transition (EMT) process by regulating corresponding adhesion molecules and transcription factors (i.e., E-cadherin, N-cadherin, Cadherin 6B, Slug and Msx1). Finally, through bioinformatic analysis and cellular thermal shift assay, we identified the AKR1B1 gene as a potential target. In summary, these findings suggest that baicalin could be used as a therapeutic agent for high glucose-induced craniofacial skeleton malformation.

Imaging the Iodine Sorption-Induced Synchronous Skeleton-Pore Interactions of Single Covalent Organic Framework Particles.

Covalent organic frameworks (COFs) are promising iodine adsorbents. For improved performances, it is critical and essential to fundamentally understand the underlying mechanism. Here, using the operando dark-field optical microscopy (DFM) imaging technique, the observation of an extraordinary structure shrinkage of 2D triphenylbenzene (TPB)-dimethoxyterephthaldehyde (DMTP)-COF upon the adsorption of I2 vapor at the single-particle resolution is reported. Combining single-particle DFM imaging with other experimental and theoretical methods, it is revealed that the shrinkage mechanism of the TPB-DMTP-COF is attributed to the I2 sorption-induced synchronous skeleton-pore interactions. The redox reaction of I2 and TPB-DMTP-COF yields some cationic skeletons and I3 - species, which triggers the multi-directional halogen-bonding interactions of I2 and I3 - as well as strong cation-π interactions between neutral and cationic skeletons, accompanying the synchronous in-plane skeleton shrinking in the xy plane and compact out-of-plane layer packing in the z-direction. This understanding of the synchronous action between the skeleton and pore breaks the perspective on the structure robustness of 2D COFs with excellent stability during the I2 uptake, which offers pivotal guidance for the rational design and creation of advanced microporous adsorbents.

Multi-Modal Enhancement Transformer Network for Skeleton-Based Human Interaction Recognition.

Skeleton-based human interaction recognition is a challenging task in the field of vision and image processing. Graph Convolutional Networks (GCNs) achieved remarkable performance by modeling the human skeleton as a topology. However, existing GCN-based methods have two problems: (1) Existing frameworks cannot effectively take advantage of the complementary features of different skeletal modalities. There is no information transfer channel between various specific modalities. (2) Limited by the structure of the skeleton topology, it is hard to capture and learn the information about two-person interactions. To solve these problems, inspired by the human visual neural network, we propose a multi-modal enhancement transformer (ME-Former) network for skeleton-based human interaction recognition. ME-Former includes a multi-modal enhancement module (ME) and a context progressive fusion block (CPF). More specifically, each ME module consists of a multi-head cross-modal attention block (MH-CA) and a two-person hypergraph self-attention block (TH-SA), which are responsible for enhancing the skeleton features of a specific modality from other skeletal modalities and modeling spatial dependencies between joints using the specific modality, respectively. In addition, we propose a two-person skeleton topology and a two-person hypergraph representation. The TH-SA block can embed their structural information into the self-attention to better learn two-person interaction. The CPF block is capable of progressively transforming the features of different skeletal modalities from low-level features to higher-order global contexts, making the enhancement process more efficient. Extensive experiments on benchmark NTU-RGB+D 60 and NTU-RGB+D 120 datasets consistently verify the effectiveness of our proposed ME-Former by outperforming state-of-the-art methods.

PAR-Net: An Enhanced Dual-Stream CNN-ESN Architecture for Human Physical Activity Recognition.

Physical exercise affects many facets of life, including mental health, social interaction, physical fitness, and illness prevention, among many others. Therefore, several AI-driven techniques have been developed in the literature to recognize human physical activities. However, these techniques fail to adequately learn the temporal and spatial features of the data patterns. Additionally, these techniques are unable to fully comprehend complex activity patterns over different periods, emphasizing the need for enhanced architectures to further increase accuracy by learning spatiotemporal dependencies in the data individually. Therefore, in this work, we develop an attention-enhanced dual-stream network (PAR-Net) for physical activity recognition with the ability to extract both spatial and temporal features simultaneously. The PAR-Net integrates convolutional neural networks (CNNs) and echo state networks (ESNs), followed by a self-attention mechanism for optimal feature selection. The dual-stream feature extraction mechanism enables the PAR-Net to learn spatiotemporal dependencies from actual data. Furthermore, the incorporation of a self-attention mechanism makes a substantial contribution by facilitating targeted attention on significant features, hence enhancing the identification of nuanced activity patterns. The PAR-Net was evaluated on two benchmark physical activity recognition datasets and achieved higher performance by surpassing the baselines comparatively. Additionally, a thorough ablation study was conducted to determine the best optimal model for human physical activity recognition.

From Movements to Metrics: Evaluating Explainable AI Methods in Skeleton-Based Human Activity Recognition.

The advancement of deep learning in human activity recognition (HAR) using 3D skeleton data is critical for applications in healthcare, security, sports, and human-computer interaction. This paper tackles a well-known gap in the field, which is the lack of testing in the applicability and reliability of XAI evaluation metrics in the skeleton-based HAR domain. We have tested established XAI metrics, namely faithfulness and stability on Class Activation Mapping (CAM) and Gradient-weighted Class Activation Mapping (Grad-CAM) to address this problem. This study introduces a perturbation method that produces variations within the error tolerance of motion sensor tracking, ensuring the resultant skeletal data points remain within the plausible output range of human movement as captured by the tracking device. We used the NTU RGB+D 60 dataset and the EfficientGCN architecture for HAR model training and testing. The evaluation involved systematically perturbing the 3D skeleton data by applying controlled displacements at different magnitudes to assess the impact on XAI metric performance across multiple action classes. Our findings reveal that faithfulness may not consistently serve as a reliable metric across all classes for the EfficientGCN model, indicating its limited applicability in certain contexts. In contrast, stability proves to be a more robust metric, showing dependability across different perturbation magnitudes. Additionally, CAM and Grad-CAM yielded almost identical explanations, leading to closely similar metric outcomes. This suggests a need for the exploration of additional metrics and the application of more diverse XAI methods to broaden the understanding and effectiveness of XAI in skeleton-based HAR.