Multiple-target tracking in human and machine vision

Kamkar, Shiva; Ghezloo, Fatemeh; Moghaddam, Hamid Abrishami; Borji, Ali; Lashgari, Reza

doi:10.1371/journal.pcbi.1007698

Cited by 18 publications

(7 citation statements)

References 116 publications

(174 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…At present, it has been widely used in many fields of production and life, and the future application prospect is even broader. Obviously, the accurate tracking or precise positioning of the target has an immeasurable impact on national security, social development, and people's life stability [6][7][8][9][10].…”

Section: Introductionmentioning

confidence: 99%

Visual Object Tracking Based on Deep Neural Network

Diao

Sun

2022

Mathematical Problems in Engineering

View full text Add to dashboard Cite

Computer vision systems cannot function without visual target tracking. Intelligent video monitoring, medical treatment, human-computer interaction, and traffic management all stand to benefit greatly from this technology. Although many new algorithms and methods emerge every year, the reality is complex. Targets are often disturbed by factors such as occlusion, illumination changes, deformation, and rapid motion. Solving these problems has also become the main task of visual target tracking researchers. As with the development for deep neural networks and attention mechanisms, object-tracking methods with deep learning show great research potential. This paper analyzes the abovementioned difficult factors, uses the tracking framework based on deep learning, and combines the attention mechanism model to accurately model the target, aiming to improve tracking algorithm. In this work, twin network tracking strategy with dual self-attention is designed. A dual self-attention mechanism is used to enhance feature representation of the target from the standpoint of space and channel, with the goal of addressing target deformation and other problems. In addition, adaptive weights and residual connections are used to enable adaptive attention feature selection. A Siamese tracking network is used in conjunction with the proposed dual self-attention technique. Massive experimental results show our proposed method improves tracking performance, and tracking strategy achieves an excellent tracking effect.

show abstract

Section: Introductionmentioning

confidence: 99%

Visual Object Tracking Based on Deep Neural Network

Diao

Sun

2022

Mathematical Problems in Engineering

View full text Add to dashboard Cite

show abstract

“…This can be done using OpenPose [22], which identifies a skeleton of 18 points at various key points that make up a human and will use those points to determine the action of the human through the distances and angles between the points. Using a multiple-target tracking (MTT) method as described in [23], each human in the frame will receive a label and their actions will be assigned to them. A report will be generated detailing each human's action and the time taken per action, and this information can be used to determine whether a given set of actions is repetitive and can be deemed as unsafe.…”

Section: Proposed Approachmentioning

confidence: 99%

An Industrial Workplace Alerting and Monitoring Platform to Prevent Workplace Injury and Accidents

Adhikesaven¹

2022

Preprint

View full text Add to dashboard Cite

Workplace accidents are a critical problem that causes many deaths, injuries, and financial losses. Climate change has a severe impact on industrial workers, partially caused by global warming. To reduce such casualties, it is important to proactively find unsafe environments where injuries could occur by detecting the use of personal protective equipment (PPE) and identifying unsafe activities. Thus, we propose an industrial workplace alerting and monitoring platform to detect PPE use and classify unsafe activity in group settings involving multiple humans and objects over a long period of time. Our proposed method is the first to analyze prolonged actions involving multiple people or objects. It benefits from combining pose estimation with PPE detection in one platform. Additionally, we propose the first open source annotated data set with video data from industrial workplaces annotated with the action classifications and detected PPE. The proposed system can be implemented within the surveillance cameras already present in industrial settings, making it a practical and effective solution. Problem and MotivationEvery year, almost two million people die from work-related accidents or diseases [1], and another 317 million suffer from work-related injuries [2]. The economic cost of these illnesses, injuries, and deaths is nearly three trillion dollars annually [3]. Beyond the direct toll on affected workers and medical expenses, workplace injuries lead to loss of productivity for all involved in the accident, property damage, and ultimately worsening morale among all workers [4]. Accounting for 39 percent of work related injuries, overexertion is the most common cause of these injuries [5]. Overexertion consists of lifting, lower, and other repetitive actions, which are all ergonomic issues. Another 31 percent of workplace injuries comes from contact with other objects [5], making the use of personal protective equipment (PPE) critical.Climate change has a strong impact on worker safety and health. Industrial workers are often most adversely affected by climate change and are exposed to physically demanding work for a longer period of time [6,7]. Both outdoor and indoor factory workers face the impacts of increased heat [6,7], which is often worsened by the lack of air conditioning and ventilation in many factories [8]. As global warming increases due to climate change, the impacts of extreme heat lead to many workplace injuries by lowering worker concentration and increasing fatigue [9]. For example, extreme heat has caused more than 20 thousands additional workplace injuries annually in California alone [9]. Overall, environmental heat causes over 170 thousand work-related injuries in the United States annually, making it the third highest cause of worker injuries [10]. It also contributes to as much as 2,000 deaths annually, possibly making heat exposure the leading cause of worker deaths [10]. Given the severe impact of climate change on workers, it is important to improve workplace safety and protect workers ...

show abstract

“…These networks can simulate the processes used by the human brain to identify, store, or process information. However, according to the research findings of [11], ANNs are fundamentally different from biological neural networks (BNNs) in many ways. For instance, cortical neurons in BNN and artificial neurons in ANN communicate to each other differently.…”

Section: Related Workmentioning

confidence: 99%

Brain Strategy Algorithm for Multiple Object Tracking Based on Merging Semantic Attributes and Appearance Features

Diab

Elhosseini

Elsayed

et al. 2021

Sensors

View full text Add to dashboard Cite

The human brain can effortlessly perform vision processes using the visual system, which helps solve multi-object tracking (MOT) problems. However, few algorithms simulate human strategies for solving MOT. Therefore, devising a method that simulates human activity in vision has become a good choice for improving MOT results, especially occlusion. Eight brain strategies have been studied from a cognitive perspective and imitated to build a novel algorithm. Two of these strategies gave our algorithm novel and outstanding results, rescuing saccades and stimulus attributes. First, rescue saccades were imitated by detecting the occlusion state in each frame, representing the critical situation that the human brain saccades toward. Then, stimulus attributes were mimicked by using semantic attributes to reidentify the person in these occlusion states. Our algorithm favourably performs on the MOT17 dataset compared to state-of-the-art trackers. In addition, we created a new dataset of 40,000 images, 190,000 annotations and 4 classes to train the detection model to detect occlusion and semantic attributes. The experimental results demonstrate that our new dataset achieves an outstanding performance on the scaled YOLOv4 detection model by achieving a 0.89 mAP 0.5.

show abstract

Multiple-target tracking in human and machine vision

Cited by 18 publications

References 116 publications

Visual Object Tracking Based on Deep Neural Network

Visual Object Tracking Based on Deep Neural Network

An Industrial Workplace Alerting and Monitoring Platform to Prevent Workplace Injury and Accidents

Brain Strategy Algorithm for Multiple Object Tracking Based on Merging Semantic Attributes and Appearance Features

Contact Info

Product

Resources

About