People Tracking Research Articles

Top view multiple people tracking by detection using deep SORT and YOLOv3 with transfer learning: within 5G infrastructure

Nowadays, 5G profoundly impacts video surveillance and monitoring services by processing video streams at high-speed with high-reliability, high bandwidth, and secure network connectivity. It also enhances artificial intelligence, machine learning, and deep learning techniques, which require intense processing to deliver near-real-time solutions. In video surveillance, person tracking is a crucial task due to the deformable nature of the human body, various environmental components such as occlusion, illumination, and background conditions, specifically, from a top view perspective where the person’s visual appearance is significantly different from a frontal or side view. In this work, multiple people tracking framework is presented, which uses 5G infrastructure. A top view perspective is used, which offers broad coverage of the scene or field of view. To perform a person tracking deep learning-based tracking by detection framework is proposed, which includes detection by YOLOv3 and tracking by Deep SORT algorithm. Although the model is pre-trained using the frontal view images, even then, it gives good detection results. In order to further enhance the accuracy of the detection model, the transfer learning approach is adopted. In this way, a detection model takes advantage of a pre-trained model appended with an additional trained layer using top view data set. To evaluate the performance, experiments are carried out on different top view video sequences. Experimental results reveal that transfer learning improves the overall performance, detection accuracy, and reduces false positives. The deep learning detection model YOLOv3 achieves detection accuracy of 92% with a pre-trained model without transfer learning and 95% with transfer learning. The tracking algorithm Deep SORT also achieves excellent results with a tracking accuracy of 96%.

Surveilling the marginalised: How manual, embodied and territorialised surveillance persists in the age of ‘dataveillance’

The emergence of new forms of data-driven surveillance – often referred to as ‘dataveillance’ – is reshaping how marginalised social groups are governed. It is generally thought that dataveillance replaces the manual monitoring of specific individuals and spaces with automated monitoring of disembodied and deterritorialised populations. This article challenges this view. Drawing on an ethnographic study of surveillance and homelessness governance in Brisbane, Australia, we argue that embodied surveillance persists in the age of dataveillance due to its capacity to address certain ‘power/knowledge challenges’ associated with the governance of marginalised social groups. We show how the manual monitoring of individuals and spaces is central to how governing actors keep track of people experiencing unsheltered homelessness, and prevent them falling into ‘surveillance gaps’ arising from the extreme social and spatial marginalisation that homelessness entails. We also show how these practices are experienced as a ‘mixed blessing’ by people experiencing homelessness, as they have the capacity to result in both punitive targeting, as well as protection and housing-focused support.

Investigating the Impact of Spatial Configuration on Users' Behaviour in Shopping Malls Case of Bab-Ezzouar Shopping Mall in Algiers

Recently, Shopping malls are introduced in Algerian cities as new commercial structures instead of traditional markets; they become a part of the daily life allowing people to have a new shopping experience. This paper aims to explore the influence of visual accessibility and spatial configuration on the navigation of individuals in a shopping mall through the analysis of Bab-Ezzouar shopping centre. The analysis of spatial and visual patterns adopts space syntax techniques and methods, a survey analysis based on people tracking and a questionnaire is applied to collect data on the shopping centre visitors’ behaviour. The results show that the perception of shopping spaces differs from familiar to unfamiliar visitors, familiar visitors are guided by their shopping habit more than spatial patterns and unfamiliar visitors are influenced by visual patterns of space more than accessibility. Visitors tend to choose the most open spaces that offer maximum visual accessibility more than physical accessibility, they also prefer walking in a straight-line avoiding change of direction. These features can guide designers in their process for better understanding of shopping space.

Convergent evidence for a theory of rapid, automatic, and accurate sex ratio tracking

It is presumed that people track the sex ratios in their environment (the number of males relative to number of females) in order to adaptively adjust their decisions and behaviors, but this actual tracking ability has not been established. The relevance of sex ratio information, drawn from evolutionary biology and studies of human relationship decision making, is integrated here with memory research (on frequency encoding), perception research (on ensemble coding), and neuroscience research. A series of four experiments provide empirical results to help fill research gaps and facilitate this theoretical integration. In particular, these studies connect details from memory research on relatively automatic frequency encoding of both items and categories, perception research on summary statistics from ensemble coding, and theoretical ideas about the function of these abilities (specifically applied to human sex ratios based on faces) from social and evolutionary approaches. Collectively this research demonstrates an evolved psychological mechanism for functional, fast, and relatively automatic human abilities to track experienced sex ratios in the social world. This sex ratio information is theorized to underpin documented facultative adjustments in relationship dynamics as well as perceptions of social group characteristics. This integrative approach highlights how the coding, memory, and judgments about population sex ratios can both account for a number of existing findings and point towards key further research.

A New Dataset of People Flow in an Industrial Site with UWB and Motion Capture Systems.

Improving performance and safety conditions in industrial sites remains a key objective for most companies. Currently, the main goal is to be able to dynamically locate both people and goods on the site. Security and access regulation to restricted areas are often ensured by doors or badge barriers and those have several issues when faced with people being in places they are not supposed to be in or even tools of objects being used incorrectly. In addition to this, a growing use of new devices requires precise information about their location in the environment such as mobile robots or drones. Therefore, it is becoming essential to have the tools to dynamically manage these flows of people and goods. Ultra-wide-band and motion capture solutions will be used to quickly identify people who may be in unauthorized areas or performing tasks which they have been uninstructed to do. In addition to the dynamic tracking of people, this also overcomes some issues associated with moving objects or tools around the production floor. We offer a new set of data that provides precise information on worker movement. This dataset can be used to develop new metrics regarding worker efficiency and safety.

Self-tracking behaviour in physical activity: a systematic review of drivers and outcomes of fitness tracking

ABSTRACT Advances in technologies (e.g. smartphones, wearables) have resulted in the concept of ‘self-tracking’, and the use of self-tracking technologies in physical activity (i.e. fitness tracking) is on the rise. For example, many people track and monitor their fitness-related metrics (e.g. steps walked, distance ran, and calories burned) to change their behaviours or keep themselves active. Despite the widespread application of self-tracking in fitness, relatively little is known about its drivers and outcomes. To address this gap, the current paper provides an overview of the literature (empirical papers) on self-tracking with a focus on the drivers and outcomes of fitness tracking behaviour and offers four important contributions. First, it identifies 19 drivers of fitness tracking technology usage. Second, it discusses four main outcomes of fitness tracking behaviour. Third, by drawing on the existing studies conducted across various fitness tracking technologies (e.g. fitness trackers, apps) and user groups (e.g. patients, seniors, and females), it provides valuable insights that can be generalisable to other settings (e.g. other types of users and fitness tracking products). Finally, the current paper provides important practical implications and addresses avenues for future research.

Caracterización pedagógica e inclusión social: un estudio de los soportes que brindan las redes personales de estudiantes en el bachillerato popular Furilofche, Argentina.

En ocasiones, las prácticas escolares se planifican de espalda al contexto cotidiano de estudiantes que enfrentan adversidades, trayectorias escolares con fracasos y estigmatizaciones. De este modo, quedan invisibilizadas historias y particularidades de los sujetos de aprendizaje que influyen en la apropiación de un conocimiento. El objetivo de este artículo es utilizar el Análisis de Redes Sociales para estudiar las redes personales de jóvenes y adultos que asisten a un Bachillerato Popular en la ciudad de Bariloche, Argentina. En acuerdo con los actores educativos de la institución se analizaron: 1) los tipos de apoyo que brindan sus redes personales dividiéndolos en apoyo emocional, tangible, informacional y axiológico; 2) los núcleos de esas redes a partir del rastreo de las personas con las que se discuten temas importantes y 3) los asuntos tratados en esos núcleos. La metodología implicó la realización de una encuesta semiestructurada que se focalizó en los estudiantes que asisten al Bachillerato y entrevistas a referentes claves. Posteriormente, se trabajó con el Software Egonet, para la graficación y análisis de las redes personales. Como resultado se observa que este tipo de estudios contribuyen a interpretar la potencialidad de ciertas estrategias didácticas sobre otras y a comprender cuáles son las diferencias existentes entre las relaciones personales fuera y dentro de los contextos escolares.

Using a Rotating 3D LiDAR on a Mobile Robot for Estimation of Person's Body Angle and Gender.

We studied the use of a rotating multi-layer 3D Light Detection And Ranging (LiDAR) sensor (specifically the Velodyne HDL-32E) mounted on a social robot for the estimation of features of people around the robot. While LiDARs are often used for robot self-localization and people tracking, we were interested in the possibility of using them to estimate the people’s features (states or attributes), which are important in human–robot interaction. In particular, we tested the estimation of the person’s body orientation and their gender. As collecting data in the real world and labeling them is laborious and time consuming, we also looked into other ways for obtaining data for training the estimators: using simulations, or using LiDAR data collected in the lab. We trained convolutional neural network-based estimators and tested their performance on actual LiDAR measurements of people in a public space. The results show that with a rotating 3D LiDAR a usable estimate of the body angle can indeed be achieved (mean absolute error 33.5), and that using simulated data for training the estimators is effective. For estimating gender, the results are satisfactory (accuracy above 80%) when the person is close enough; however, simulated data do not work well and training needs to be done on actual people measurements.

Perspectives on usability and accessibility of an autonomous humanoid robot living with elderly people

Purpose The aim of this study was to assess how cohabitation with a social robot for seven days night and day is perceived by elderly potential users and to suggest technological improvements for future home integration. Methods Fifteen participants were invited to interact with Pepper®, 24 h a day and 7 days a week. Perceived utility assessment was achieved using a dedicated questionnaire. Usability assessment was carried out by measuring satisfaction with the human-robot interaction and listing malfunctions. Results Seventy-eight tasks out of 115 tasks were reported respectively as very high and high. On day 1, satisfaction was at its higher level with a mean of 6.8. On day 8, it maintained at 6.0 despite technical malfunctions and failures, listed with details in this paper. Acceptability was positive. Pepper was mainly considered as an assistant and a link with family and friends and never as an enemy or intruder. Pepper has proven to be potentially able to meet five needs: making the person’s environment safe by alerting contact persons, reminding the user of health-related tasks, enabling “doubt removal”, maintaining social relationships and following the user. Conclusions Pepper is more than a “demonstration” device classified as level 6 of the nine-point Technology Readiness Level (TRL) scale for the hard and software and level 5 in behavioural terms. Once the technological locks and malfunctions have been overcome, Pepper should be applicable in the home, allowing users to test it in a more familiar environment. Implications for Rehabilitation Pepper, a humanoid robot, has been perceived as rich in potentialities that can help and secure a fragile person. Following this experience of cohabitation with a humanoid robot, the participants were unanimous in thinking that assistive robotics is the future. To be useful, the robot’s people tracking function must inevitably be improved. Pepper, the first humanoid robot ever marketed, is more than just a “demonstration”device, classified at level 6 of Technological Readiness Levels (TRL) for supporting hardware and software and level 5 in terms of behavior.

Jointly detecting and multiple people tracking by semantic and scene information

In this paper, we propose a new method for online multiple people tracking, which combines the detection process and the single object tracking process, and establishes the interactions between them. The detector detects objects in the still images which ignores the sequential information. Meantime, the single object tracker does not use the category semantic information during tracking. To take both the sequential and semantic information into account, we exchange information among the detector and the trackers. More specifically, the trackers deliver sequential information to the detector by providing the detector with the extra proposals. The detector supplements each tracker with the robust semantic information by using bounding box regression to modify the tracking result. Besides, the interactions also happen among the trackers through the occlusion speculation, the perspective model interpretation and the trajectory merging process. The experimental results demonstrate that the proposed algorithm performs favorably against the state-of-the-art MOT methods.

Simulating Signal Aberration and Ranging Error for Ultrasonic Indoor Positioning.

Increasing efforts toward the development of positioning techniques testify the growing interest for indoor position-based applications and services. Many applications require accurate indoor positioning or tracking of people and assets, and some market sectors are starting a rapid growth of products based on these technologies. Ultrasonic systems have already been demonstrating their effectiveness and to possess the desired positioning accuracy and refresh rates. In this work, it is shown that a typical signal used in ultrasonic positioning systems to estimate the range between the target and reference points—namely, the linear chirp—due to the effects of acoustic diffraction, in some cases, undergoes a shape aberration, depending on the shape and size of the transducer and on the angle under which the transducer is seen by the receiver. In the presence of such signal shape aberrations, even one of the most robust ranging techniques, which is based on cross-correlation, provides results affected by a much greater error than expected. Numerical simulations are carried out for a typical ultrasonic chirp, ultrasonic emitter, and range technique based on cross-correlation and for a typical office room, obtained using the academic acoustic simulation software Field II. Spatial distributions of the ranging error are provided, clearly showing the favorable low error regions. The work demonstrates that particular attention must be paid to the design of the acoustic section of the ultrasonic positioning systems, considering both the shape and size of the ultrasonic emitters and the shape of the acoustic signal used.

Leveraging Polarization of WiFi Signals to Simultaneously Track Multiple People

This paper presents WiPolar, an approach that simultaneously tracks multiple people using commodity WiFi devices. While two recent papers have also demonstrated multi-person tracking using commodity devices, they either require the people to continuously keep moving without stopping, and/or require the number of people to be input manually, and/or keep the WiFi devices from performing their primary function of data communication. Motivated by the increasing availability of polarized antennas on modern WiFi devices, WiPolar leverages signal polarization to perform accurate multi-person tracking using commodity devices while addressing the three limitations of prior work mentioned above. The key insight that WiPolar is based on is that different people expose different instantaneous horizontal and vertical radar cross-sections to WiFi transmitters due to differences in their physiques and orientations with respect to the transmitter. This enables WiPolar to accurately separate the multipaths reflected from different people, which, in turn, allows it to track them simultaneously. To the best of our knowledge, this is the first work that leverages polarization of WiFi signals to localize and track people. We implement WiPolar using commodity WiFi devices and extensively evaluate it for tracking up to five people in three different environments. Our results show that WiPolar achieved a median tracking error of just 56cm across all experiments. It also accurately tracks people even when they were not moving. WiPolar achieved a median tracking error of 74cm for people that were either stationary or just taking a small pause.

Human Gait Tracking for Normal People and Walker Users Using a 2D LiDAR

This paper proposes a human gait tracking method using a low scanning rate (10Hz) 360 degree 2D LiDAR. The sensor is placed at a height of 0.3m above the ground. Since the human gait tracking is important not only for normal people but also for patients who need to use walking-aid such as walker, we propose a tracking method for both normal people and walker users. The left and right legs are separated and tracked even with nonlinear walking path. To overcome the occlusions between legs when one leg might be hidden from the sensor, a spline-based smooth and interpolation method is applied to estimate the legs trajectories. An algorithm is also proposed to remove the walker legs data in case of walker users. The performance of proposed tracking method is investigated in different walking speed levels and two positions of LiDAR (parallel and horizontal to walking direction). The experimental results verify the robustness of our proposed method. The horizontal position of LiDAR gives better estimation than the parallel position. The RMSE is less than 2.0cm for walking step length estimation and is less than 1.4cm for walking distance estimation. The tracking range of our system is up to 10m with horizontal placement of LiDAR.

Research on the Creation of Vibrant Public Space Suitable for People of all Ages in the Residential Block Based on Superposition Analysis of Behaviour Track——Examination of North Street of Sisheng Ancestral Temple in Chengdu City as an Example

As the dynamic carrier of different age groups, block public space is closely related to its space. This study takes the public space of the north street block of Sisheng ancestral temple as the research object. Through the field research of space observation, behaviour tracking and interviews of people of all ages, this paper studies the public space composition and characteristics of the north street block of Sisheng ancestral temple, analyzes the behaviour types and characteristics of different age groups, extracts the space elements, further analyzes the degree between space and behaviour correlation. On this basis, according to the space use needs of people of all ages, this paper puts forward strategies for creating vitality of public space for people of mixed ages in the residential blocks.

Identifier of human emotions based on convolutional neural network for assistant robot

This paper proposes a solution for the problem of continuous prediction in real-time of the emotional state of a human user from the identification of characteristics in facial expressions. In robots whose main task is the care of people (children, sick or elderly people) is important to maintain a close relationship man-machine, anld a rapid response of the robot to the actions of the person under care. We propose to increase the level of intimacy of the robot, and its response to specific situations of the user, identifying in real time the emotion reflected by the person's face. This solution is integrated with algorithms of the research group related to the tracking of people for use on an assistant robot. The strategy used involves two stages of processing, the first involves the detection of faces using HOG and linear SVM, while the second identifies the emotion in the face using a CNN. The strategy was completely tested in the laboratory on our robotic platform, demonstrating high performance with low resource consumption. Through various controlled laboratory tests with different people, which forced a certain emotion on their faces, the scheme was able to identify the emotions with a success rate of 92%.

Other people’s gaze encoded as implied motion in the human brain

Keeping track of other people's gaze is an essential task in social cognition and key for successfully reading other people's intentions and beliefs (theory of mind). Recent behavioral evidence suggests that we construct an implicit model of other people's gaze, which may incorporate physically incoherent attributes such as a construct of force-carrying beams that emanate from the eyes. Here, we used functional magnetic resonance imaging and multivoxel pattern analysis to test the prediction that the brain encodes gaze as implied motion streaming from an agent toward a gazed-upon object. We found that a classifier, trained to discriminate the direction of visual motion, significantly decoded the gaze direction in static images depicting a sighted face, but not a blindfolded one, from brain activity patterns in the human motion-sensitive middle temporal complex (MT+) and temporo-parietal junction (TPJ). Our results demonstrate a link between the visual motion system and social brain mechanisms, in which the TPJ, a key node in theory of mind, works in concert with MT+ to encode gaze as implied motion. This model may be a fundamental aspect of social cognition that allows us to efficiently connect agents with the objects of their attention. It is as if the brain draws a quick visual sketch with moving arrows to help keep track of who is attending to what. This implicit, fluid-flow model of other people's gaze may help explain culturally universal myths about the mind as an energy-like, flowing essence.

Consumer Behavior Analysis in the Offline Retail Stores based on convolutional neural network

Pedestrian attributes recognizing (PAR) is an important task in computer vision area due to it plays an important role in video surveillance. On the other hand, pedestrian visual attributes are treated as middle-level semantic features which can provide calibration information for high- level human related visual tasks in order to improve the discriminative ability of the models, such as pedestrian detection, people tracking, person re-identification, action recognition and scene understanding. However, the most study of PAR is based on single person. In this work, I implements a multiple pedestrian attributes recognition model based on the offline retail scenes. This model combines object detection and multitask classification techniques and can be trained end-to-end directly for back propagation. This paper also demonstrates the performance of the final model through a series of experiment.

Semi-online Multi-people Tracking by Re-identification

In this paper, we propose a novel semi-online approach to tracking multiple people. In contrast to conventional offline approaches that take the whole image sequence as input, our semi-online approach tracks people in a frame-by-frame manner by exploring the time, space and multi-camera relationship of detection hypotheses in the near future frames. We cast the multi-people tracking task as a re-identification problem, and explicitly account for objects’ appearance changes and longer-term associations. We model our approach using a Multi-Label Markov Random Field, and introduce a fast $$\alpha $$-expansion algorithm to solve it efficiently. To our best knowledge, this is the first semi-online approach achieved by re-identification. It yields very promising tracking results especially in challenging cases, such as scenarios of the crowded streets where pedestrians frequently occlude each other, scenes captured with moving cameras where objects may disappear and reappear randomly, and videos under changing illuminations wherein the appearances of objects are influenced.

UWB/BLE Tracking System for Elderly People Monitoring.

Localization systems are the source of data that allows to evaluate elderly person’s behaviour, to draw conclusions concerning his or her health status and wellbeing, and to detect emergency situations. The article contains a description of a system intended for elderly people tracking. Two novel solutions have been implemented in the system: a hybrid localization algorithm and a method for wireless anchor nodes synchronization. The algorithm fuses results of time difference of arrival and received signal strength measurements in ultrawideband (UWB) and Bluetooth Low Energy (BLE) radio interfaces, respectively. The system allows to change the intensity of UWB packets transmission to adapt localization accuracy and energy usage to current needs and applications. In order to simplify the system installation, communication between elements of the system infrastructure instead of wire interfaces is performed over wireless ones. The new wireless synchronization method proposed in the article consists in retransmission of UWB synchronization packets by selected anchor nodes. It allows for extension of the system coverage, which is limited by the short range of UWB transmission. The proposed solution was experimentally verified. The synchronization method was tested in a laboratory, and the whole system’s performance was investigated in a typical flat. Exemplary results of the tests performed with older adult participation in their own homes are also included.

Respiration Tracking for People Counting and Recognition

Wireless detection of respiration rates is crucial for many applications. Most of the state-of-the-art solutions estimate breathing rates with the prior knowledge of crowd numbers as well as assuming the distinct breathing rates of different users, which is neither natural nor realistic. However, few of them can leverage the estimated breathing rates to recognize human subjects (also known as identity matching). In this article, using the channel state information (CSI) of a single pair of commercial WiFi devices, a novel system is proposed to continuously track the breathing rates of multiple persons without such impractical assumptions. The proposed solution includes an adaptive subcarrier combination method that boosts the signal-to-noise ratio (SNR) of breathing signals, and iterative dynamic programming and a trace concatenating algorithm that continuously tracks the breathing rates of multiple users. By leveraging both the spectrum and time diversity of the CSI, our system can correctly extract the breathing rate traces even if some of them merge together for a short time period. Furthermore, by utilizing the breathing traces obtained, our system can do people counting and recognition simultaneously. Extensive experiments are conducted in two environments (an on-campus lab and a car). The results show that 86% of average accuracy can be achieved for people counting up to four people for both cases. For 97.9% out of all the testing cases, the absolute error of crowd number estimates is within 1. The system achieves an average accuracy of 85.78% for people recognition in a smart home case.