Human Gait Recognition Based on Frame-by-Frame Gait Energy Images and Convolutional Long Short-Term Memory

Computer Vision-Based Gait Recognition on the Edge: A Survey on Feature Representations, Models, and Architectures

Computer vision-based gait recognition (CVGR) is a technology that has gained considerable attention in recent years due to its non-invasive, unobtrusive, and difficult-to-conceal nature. Beyond its applications in biometrics, CVGR holds significant potential for healthcare and human-computer interaction. Current CVGR systems often transmit collected data to a cloud server for machine learning-based gait pattern recognition. While effective, this cloud-centric approach can result in increased system response times. Alternatively, the emerging paradigm of edge computing, which involves moving computational processes to local devices, offers the potential to reduce latency, enable real-time surveillance, and eliminate reliance on internet connectivity. Furthermore, recent advancements in low-cost, compact microcomputers capable of handling complex inference tasks (e.g., Jetson Nano Orin, Jetson Xavier NX, and Khadas VIM4) have created exciting opportunities for deploying CVGR systems at the edge. This paper reports the state of the art in gait data acquisition modalities, feature representations, models, and architectures for CVGR systems suitable for edge computing. Additionally, this paper addresses the general limitations and highlights new avenues for future research in the promising intersection of CVGR and edge computing.

A comprehensive review of gait analysis using deep learning approaches in criminal investigation

Despite the growing worries expressed by privacy supporters about the extensive adoption of gait biometrics, research in this field has been moving forward swiftly. Deep learning, a powerful technology that enables computers to learn from data, has found its way into criminal investigations involving gait. In this survey, the literature of gait analysis concerning criminal investigation is discussed with a comprehensive overview of developments in gait analysis with deep neural networks. Firstly, terminologies and factors regarding human gait with scenarios related to crime are discussed. Subsequently, the areas and domains corresponding to criminal investigation that can be tackled by gait analysis are discussed. Also, deep learning methods for gait analysis and how they can be applied in criminal investigations are presented. Then, gait analysis techniques and approaches using deep learning methods including currently available datasets are mentioned. Moreover, crime-related video datasets are presented with literature on deep learning-based anomaly detection with gait human poses. Finally, challenges regarding gait analysis in criminal investigations are presented with open research issues.

Abnormal Behavior Recognition Based on 3D Dense Connections

Abnormal behavior recognition is an important technology used to detect and identify activities or events that deviate from normal behavior patterns. It has wide applications in various fields such as network security, financial fraud detection, and video surveillance. In recent years, Deep Convolution Networks (ConvNets) have been widely applied in abnormal behavior recognition algorithms and have achieved significant results. However, existing abnormal behavior detection algorithms mainly focus on improving the accuracy of the algorithms and have not explored the real-time nature of abnormal behavior recognition. This is crucial to quickly identify abnormal behavior in public places and improve urban public safety. Therefore, this paper proposes an abnormal behavior recognition algorithm based on three-dimensional (3D) dense connections. The proposed algorithm uses a multi-instance learning strategy to classify various types of abnormal behaviors, and employs dense connection modules and soft-threshold attention mechanisms to reduce the model's parameter count and enhance network computational efficiency. Finally, redundant information in the sequence is reduced by attention allocation to mitigate its negative impact on recognition results. Experimental verification shows that our method achieves a recognition accuracy of 95.61% on the UCF-crime dataset. Comparative experiments demonstrate that our model has strong performance in terms of recognition accuracy and speed.

A Novel Multi-Scaled Deep Convolutional Structure for Punctilious Human Gait Authentication

The need for non-interactive human recognition systems to ensure safe isolation between users and biometric equipment has been exposed by the COVID-19 pandemic. This study introduces a novel Multi-Scaled Deep Convolutional Structure for Punctilious Human Gait Authentication (MSDCS-PHGA). The proposed MSDCS-PHGA involves segmenting, preprocessing, and resizing silhouette images into three scales. Gait features are extracted from these multi-scale images using custom convolutional layers and fused to form an integrated feature set. This multi-scaled deep convolutional approach demonstrates its efficacy in gait recognition by significantly enhancing accuracy. The proposed convolutional neural network (CNN) architecture is assessed using three benchmark datasets: CASIA, OU-ISIR, and OU-MVLP. Moreover, the proposed model is evaluated against other pre-trained models using key performance metrics such as precision, accuracy, sensitivity, specificity, and training time. The results indicate that the proposed deep CNN model outperforms existing models focused on human gait. Notably, it achieves an accuracy of approximately 99.9% for both the CASIA and OU-ISIR datasets and 99.8% for the OU-MVLP dataset while maintaining a minimal training time of around 3 min.

Enhanced Multitask Learning for Hash Code Generation of Palmprint Biometrics

This paper presents a novel multitask learning framework for palmprint biometrics, which optimizes classification and hashing branches jointly. The classification branch within our framework facilitates the concurrent execution of three distinct tasks: identity recognition and classification of soft biometrics, encompassing gender and chirality. On the other hand, the hashing branch enables the generation of palmprint hash codes, optimizing for minimal storage as templates and efficient matching. The hashing branch derives the complementary information from these tasks by amalgamating knowledge acquired from the classification branch. This approach leads to superior overall performance compared to individual tasks in isolation. To enhance the effectiveness of multitask learning, two additional modules, an attention mechanism module and a customized gate control module, are introduced. These modules are vital in allocating higher weights to crucial channels and facilitating task-specific expert knowledge integration. Furthermore, an automatic weight adjustment module is incorporated to optimize the learning process further. This module fine-tunes the weights assigned to different tasks, improving performance. Integrating the three modules above has shown promising accuracies across various classification tasks and has notably improved authentication accuracy. The extensive experimental results validate the efficacy of our proposed framework.

Self-Supervised Learning for Electroencephalography

Decades of research have shown machine learning superiority in discovering highly nonlinear patterns embedded in electroencephalography (EEG) records compared with conventional statistical techniques. However, even the most advanced machine learning techniques require relatively large, labeled EEG repositories. EEG data collection and labeling are costly. Moreover, combining available datasets to achieve a large data volume is usually infeasible due to inconsistent experimental paradigms across trials. Self-supervised learning (SSL) solves these challenges because it enables learning from EEG records across trials with variable experimental paradigms, even when the trials explore different phenomena. It aggregates multiple EEG repositories to increase accuracy, reduce bias, and mitigate overfitting in machine learning training. In addition, SSL could be employed in situations where there is limited labeled training data, and manual labeling is costly. This article: 1) provides a brief introduction to SSL; 2) describes some SSL techniques employed in recent studies, including EEG; 3) proposes current and potential SSL techniques for future investigations in EEG studies; 4) discusses the cons and pros of different SSL techniques; and 5) proposes holistic implementation tips and potential future directions for EEG SSL practices.

A Class-Imbalance Aware and Explainable Spatio-Temporal Graph Attention Network for Neonatal Seizure Detection

Seizures are the most prevalent clinical indication of neurological disorders in neonates. In this study, a class-imbalance aware and explainable deep learning approach based on Convolutional Neural Networks (CNNs) and Graph Attention Networks (GATs) is proposed for the accurate automated detection of neonatal seizures. The proposed model integrates the temporal information of EEG signals with the spatial information on the EEG channels through the graph representation of the multi-channel EEG segments. One-dimensional CNNs are used to automatically develop a feature set that accurately represents the differences between seizure and nonseizure epochs in the time domain. By employing GAT, the attention mechanism is utilized to emphasize the critical channel pairs and information flow among brain regions. GAT coefficients were then used to empirically visualize the important regions during the seizure and nonseizure epochs, which can provide valuable insight into the location of seizures in the neonatal brain. Additionally, to tackle the severe class imbalance in the neonatal seizure dataset using under-sampling and focal loss techniques are used. Overall, the final Spatio-Temporal Graph Attention Network (ST-GAT) outperformed previous benchmarked methods with a mean AUC of 96.6% and Kappa of 0.88, demonstrating its high accuracy and potential for clinical applications.

Person Recognition Based on Deep Gait: A Survey

Gait recognition, also known as walking pattern recognition, has expressed deep interest in the computer vision and biometrics community due to its potential to identify individuals from a distance. It has attracted increasing attention due to its potential applications and non-invasive nature. Since 2014, deep learning approaches have shown promising results in gait recognition by automatically extracting features. However, recognizing gait accurately is challenging due to the covariate factors, complexity and variability of environments, and human body representations. This paper provides a comprehensive overview of the advancements made in this field along with the challenges and limitations associated with deep learning methods. For that, it initially examines the various gait datasets used in the literature review and analyzes the performance of state-of-the-art techniques. After that, a taxonomy of deep learning methods is presented to characterize and organize the research landscape in this field. Furthermore, the taxonomy highlights the basic limitations of deep learning methods in the context of gait recognition. The paper is concluded by focusing on the present challenges and suggesting several research directions to improve the performance of gait recognition in the future.

Deep Gait Recognition: A Survey

Gait recognition is an appealing biometric modality which aims to identify individuals based on the way they walk. Deep learning has reshaped the research landscape in this area since 2015 through the ability to automatically learn discriminative representations. Gait recognition methods based on deep learning now dominate the state-of-the-art in the field and have fostered real-world applications. In this paper, we present a comprehensive overview of breakthroughs and recent developments in gait recognition with deep learning, and cover broad topics including datasets, test protocols, state-of-the-art solutions, challenges, and future research directions. We first review the commonly used gait datasets along with the principles designed for evaluating them. We then propose a novel taxonomy made up of four separate dimensions namely body representation, temporal representation, feature representation, and neural architecture, to help characterize and organize the research landscape and literature in this area. Following our proposed taxonomy, a comprehensive survey of gait recognition methods using deep learning is presented with discussions on their performances, characteristics, advantages, and limitations. We conclude this survey with a discussion on current challenges and mention a number of promising directions for future research in gait recognition.

Advances in Vision-Based Gait Recognition: From Handcrafted to Deep Learning

Identifying people’s identity by using behavioral biometrics has attracted many researchers’ attention in the biometrics industry. Gait is a behavioral trait, whereby an individual is identified based on their walking style. Over the years, gait recognition has been performed by using handcrafted approaches. However, due to several covariates’ effects, the competence of the approach has been compromised. Deep learning is an emerging algorithm in the biometrics field, which has the capability to tackle the covariates and produce highly accurate results. In this paper, a comprehensive overview of the existing deep learning-based gait recognition approach is presented. In addition, a summary of the performance of the approach on different gait datasets is provided.

Uncertainty-Guided Voxel-Level Supervised Contrastive Learning for Semi-Supervised Medical Image Segmentation

Semi-supervised learning reduces overfitting and facilitates medical image segmentation by regularizing the learning of limited well-annotated data with the knowledge provided by a large amount of unlabeled data. However, there are many misuses and underutilization of data in conventional semi-supervised methods. On the one hand, the model will deviate from the empirical distribution under the training of numerous unlabeled data. On the other hand, the model treats labeled and unlabeled data differently and does not consider inter-data information. In this paper, a semi-supervised method is proposed to exploit unlabeled data to further narrow the gap between the semi-supervised model and its fully-supervised counterpart. Specifically, the architecture of the proposed method is based on the mean-teacher framework, and the uncertainty estimation module is improved to impose constraints of consistency and guide the selection of feature representation vectors. Notably, a voxel-level supervised contrastive learning module is devised to establish a contrastive relationship between feature representation vectors, whether from labeled or unlabeled data. The supervised manner ensures that the network learns the correct knowledge, and the dense contrastive relationship further extracts information from unlabeled data. The above overcomes data misuse and underutilization in semi-supervised frameworks. Moreover, it favors the feature representation with intra-class compactness and inter-class separability and gains extra performance. Extensive experimental results on the left atrium dataset from Atrial Segmentation Challenge demonstrate that the proposed method has superior performance over the state-of-the-art methods.

Human Silhouette and Skeleton Video Synthesis Through Wi-Fi signals

The increasing availability of wireless access points (APs) is leading toward human sensing applications based on Wi-Fi signals as support or alternative tools to the widespread visual sensors, where the signals enable to address well-known vision-related problems such as illumination changes or occlusions. Indeed, using image synthesis techniques to translate radio frequencies to the visible spectrum can become essential to obtain otherwise unavailable visual data. This domain-to-domain translation is feasible because both objects and people affect electromagnetic waves, causing radio and optical frequencies variations. In the literature, models capable of inferring radio-to-visual features mappings have gained momentum in the last few years since frequency changes can be observed in the radio domain through the channel state information (CSI) of Wi-Fi APs, enabling signal-based feature extraction, e.g. amplitude. On this account, this paper presents a novel two-branch generative neural network that effectively maps radio data into visual features, following a teacher-student design that exploits a cross-modality supervision strategy. The latter conditions signal-based features in the visual domain to completely replace visual data. Once trained, the proposed method synthesizes human silhouette and skeleton videos using exclusively Wi-Fi signals. The approach is evaluated on publicly available data, where it obtains remarkable results for both silhouette and skeleton videos generation, demonstrating the effectiveness of the proposed cross-modality supervision strategy.

Deep Learning and Kurtosis-Controlled, Entropy-Based Framework for Human Gait Recognition Using Video Sequences

Gait is commonly defined as the movement pattern of the limbs over a hard substrate, and it serves as a source of identification information for various computer-vision and image-understanding techniques. A variety of parameters, such as human clothing, angle shift, walking style, occlusion, and so on, have a significant impact on gait-recognition systems, making the scene quite complex to handle. In this article, we propose a system that effectively handles problems associated with viewing angle shifts and walking styles in a real-time environment. The following steps are included in the proposed novel framework: (a) real-time video capture, (b) feature extraction using transfer learning on the ResNet101 deep model, and (c) feature selection using the proposed kurtosis-controlled entropy (KcE) approach, followed by a correlation-based feature fusion step. The most discriminant features are then classified using the most advanced machine learning classifiers. The simulation process is fed by the CASIA B dataset as well as a real-time captured dataset. On selected datasets, the accuracy is 95.26% and 96.60%, respectively. When compared to several known techniques, the results show that our proposed framework outperforms them all.

Deep supervised hashing for gait retrieval

Background: Gait recognition is perceived as the most promising biometric approach for future decades especially because of its efficient applicability in surveillance systems. Due to recent growth in the use of gait biometrics across surveillance systems, the ability to rapidly search for the required data has become an emerging need. Therefore, we addressed the gait retrieval problem, which retrieves people with gaits similar to a query subject from a large-scale dataset. Methods: This paper presents the deep gait retrieval hashing (DGRH) model to address the gait retrieval problem for large-scale datasets. Our proposed method is based on a supervised hashing method with a deep convolutional network. We use the ability of the convolutional neural network (CNN) to capture the semantic gait features for feature representation and learn the compact hash codes with the compatible hash function. Therefore, our DGRH model combines gait feature learning with binary hash codes. In addition, the learning loss is designed with a classification loss function that learns to preserve similarity and a quantization loss function that controls the quality of the hash codes Results: The proposed method was evaluated against the CASIA-B, OUISIR-LP, and OUISIR-MVLP benchmark datasets and received the promising result for gait retrieval tasks. Conclusions: The end-to-end deep supervised hashing model is able to learn discriminative gait features and is efficient in terms of the storage memory and speed for gait retrieval.

Deep supervised hashing for gait retrieval

Background: Gait recognition is perceived as the most promising biometric approach for future decades especially because of its efficient applicability in surveillance systems. Due to recent growth in the use of gait biometrics across surveillance systems, the ability to rapidly search for the required data has become an emerging need. Therefore, we addressed the gait retrieval problem, which retrieves people with gaits similar to a query subject from a large-scale dataset. Methods: This paper presents the deep gait retrieval hashing (DGRH) model to address the gait retrieval problem for large-scale datasets. Our proposed method is based on a supervised hashing method with a deep convolutional network. We use the ability of the convolutional neural network (CNN) to capture the semantic gait features for feature representation and learn the compact hash codes with the compatible hash function. Therefore, our DGRH model combines gait feature learning with binary hash codes. In addition, the learning loss is designed with a classification loss function that learns to preserve similarity and a quantization loss function that controls the quality of the hash codes Results: The proposed method was evaluated against the CASIA-B, OUISIR-LP, and OUISIR-MVLP benchmark datasets and received the promising result for gait retrieval tasks. Conclusions: The end-to-end deep supervised hashing model is able to learn discriminative gait features and is efficient in terms of the storage memory and speed for gait retrieval.

Detection of Trees on Street-View Images Using a Convolutional Neural Network

Real-time detection of possible deforestation of urban landscapes is an essential task for many urban forest monitoring services. Computational methods emerge as a rapid and efficient solution to evaluate bird's-eye-view images taken by satellites, drones, or even street-view photos captured at the ground level of the urban scenery. Identifying unhealthy trees requires detecting the tree itself and its constituent parts to evaluate certain aspects that may indicate unhealthiness, being street-level images a cost-effective and feasible resource to support the fieldwork survey. This paper proposes detecting trees and their specific parts on street-view images through a Convolutional Neural Network model based on the well-known You Only Look Once network with a MobileNet as the backbone for feature extraction. Essentially, from a photo taken from the ground, the proposed method identifies trees, isolates them through their bounding boxes, identifies the crown and stem, and then estimates the height of the trees by using a specific handheld object as a reference in the images. Experiment results demonstrate the effectiveness of the proposed method.

Arbitrary Scale Super-Resolution for Medical Images

Single image super-resolution (SISR) aims to obtain a high-resolution output from one low-resolution image. Currently, deep learning-based SISR approaches have been widely discussed in medical image processing, because of their potential to achieve high-quality, high spatial resolution images without the cost of additional scans. However, most existing methods are designed for scale-specific SR tasks and are unable to generalize over magnification scales. In this paper, we propose an approach for medical image arbitrary-scale super-resolution (MIASSR), in which we couple meta-learning with generative adversarial networks (GANs) to super-resolve medical images at any scale of magnification in [Formula: see text]. Compared to state-of-the-art SISR algorithms on single-modal magnetic resonance (MR) brain images (OASIS-brains) and multi-modal MR brain images (BraTS), MIASSR achieves comparable fidelity performance and the best perceptual quality with the smallest model size. We also employ transfer learning to enable MIASSR to tackle SR tasks of new medical modalities, such as cardiac MR images (ACDC) and chest computed tomography images (COVID-CT). The source code of our work is also public. Thus, MIASSR has the potential to become a new foundational pre-/post-processing step in clinical image analysis tasks such as reconstruction, image quality enhancement, and segmentation.

Gait Recognition Using Optical Motion Capture: A Decision Fusion Based Method

Human identification based on motion capture data has received signification attentions for its wide applications in authentication and surveillance systems. The optical motion capture system (OMCS) can dynamically capture the high-precision three-dimensional locations of optical trackers that are implemented on a human body, but its potential in applications on gait recognition has not been studied in existing works. On the other hand, a typical OMCS can only support one player one time, which limits its capability and efficiency. In this paper, our goals are investigating the performance of OMCS-based gait recognition performance, and realizing gait recognition in OMCS such that it can support multiple players at the same time. We develop a gait recognition method based on decision fusion, and it includes the following four steps: feature extraction, unreliable feature calibration, classification of single motion frame, and decision fusion of multiple motion frame. We use kernel extreme learning machine (KELM) for single motion classification, and in particular we propose a reliability weighted sum (RWS) decision fusion method to combine the fuzzy decisions of the motion frames. We demonstrate the performance of the proposed method by using walking gait data collected from 76 participants, and results show that KELM significantly outperforms support vector machine (SVM) and random forest in the single motion frame classification task, and demonstrate that the proposed RWS decision fusion rule can achieve better fusion accuracy compared with conventional fusion rules. Our results also show that, with 10 motion trackers that are implemented on lower body locations, the proposed method can achieve 100% validation accuracy with less than 50 gait motion frames.

Human Gait Recognition Based on Self-Adaptive Hidden Markov Model

Human gait recognition has numerous challenges due to view angle changing, human dressing, bag carrying, and pedestrian walking speed, etc. In order to increase gait recognition accuracy under these circumstances, in this paper we propose a method for gait recognition based on a self-adaptive hidden Markov model (SAHMM). First, we present a feature extraction algorithm based on local gait energy image (LGEI) and construct an observation vector set. By using this set, we optimize parameters of the SAHMM-based method for gait recognition. Finally, the proposed method is evaluated extensively based on the CASIA Dataset B for gait recognition under various conditions such as cross view, human dressing, or bag carrying, etc. Furthermore, the generalization ability of this method is verified based on the OU-ISIR Large Population Dataset. Both experimental results show that the proposed method exhibits superior performance in comparison with those existing methods.

Using a Deep Learning Method and Data from Two-Dimensional (2D) Marker-Less Video-Based Images for Walking Speed Classification

Human body measurement data related to walking can characterize functional movement and thereby become an important tool for health assessment. Single-camera-captured two-dimensional (2D) image sequences of marker-less walking individuals might be a simple approach for estimating human body measurement data which could be used in walking speed-related health assessment. Conventional body measurement data of 2D images are dependent on body-worn garments (used as segmental markers) and are susceptible to changes in the distance between the participant and camera in indoor and outdoor settings. In this study, we propose five ratio-based body measurement data that can be extracted from 2D images and can be used to classify three walking speeds (i.e., slow, normal, and fast) using a deep learning-based bidirectional long short-term memory classification model. The results showed that average classification accuracies of 88.08% and 79.18% could be achieved in indoor and outdoor environments, respectively. Additionally, the proposed ratio-based body measurement data are independent of body-worn garments and not susceptible to changes in the distance between the walking individual and camera. As a simple but efficient technique, the proposed walking speed classification has great potential to be employed in clinics and aged care homes.

Multi-View Gait Image Generation for Cross-View Gait Recognition

Gait recognition aims to recognize persons' identities by walking styles. Gait recognition has unique advantages due to its characteristics of non-contact and long-distance compared with face and fingerprint recognition. Cross-view gait recognition is a challenge task because view variance may produce large impact on gait silhouettes. The development of deep learning has promoted cross-view gait recognition performances to a higher level. However, performances of existing deep learning-based cross-view gait recognition methods are limited by lack of gait samples under different views. In this paper, we take a Multi-view Gait Generative Adversarial Network (MvGGAN) to generate fake gait samples to extend existing gait datasets, which provides adequate gait samples for deep learning-based cross-view gait recognition methods. The proposed MvGGAN method trains a single generator for all view pairs involved in single or multiple datasets. Moreover, we perform domain alignment based on projected maximum mean discrepancy to reduce the influence of distribution divergence caused by sample generation. The experimental results on CASIA-B and OUMVLP dataset demonstrate that fake gait samples generated by the proposed MvGGAN method can improve performances of existing state-of-the-art cross-view gait recognition methods obviously on both single-dataset and cross-dataset evaluation settings.

Multiscaled Neural Autoregressive Distributed Lag: A New Empirical Mode Decomposition Model for Nonlinear Time Series Forecasting

Forecasting has always been the cornerstone of machine learning and statistics. Despite the great evolution of the time series theory, forecasters are still in the hunt for better models to make more accurate decisions. The huge advances in neural networks over the last years has led to the emergence of a new generation of effective models replacing classic econometric models. It is in this direction that we propose, in this paper, a new multiscaled Feedforward Neural Network (FNN), with the aim of forecasting multivariate time series. This new model, called Empirical Mode Decomposition (EMD)-based Neural ARDL, is inspired from the well-known Autoregressive Distributed Lag (ARDL) model being our proposal founded upon the concepts of nonlinearity, EMD-multiresolution and neural networks. These features give the model the ability to effectively capture many nonlinear patterns like the ones often present in econophysical time series, such as nonlinear trends, seasonal effects, long-range dependency, etc. The proposed algorithm can be summarized into the following four basic tasks: (i) EMD breaking-down multivariate time series into different resolution levels, (ii) feeding EMD components from the same levels into a number of feedforward neural ARDL models, (iii) from one level to the next, extrapolating the component corresponding to the response variable (scalar output) a number of steps ahead, and finally, (iv) recombining level-by-level forecasts into a single output. An optimal learning scheme is rigorously designed for efficiently training the new proposed architecture. The approach is finally tested and compared to a number of powerful benchmark models, where experiments are conducted on real-world data.