Showing papers on "Task analysis published in 2021"

A continual learning survey: Defying forgetting in classification tasks.

Abstract: Artificial neural networks thrive in solving the classification problem for a particular rigid task, acquiring knowledge through generalized learning behaviour from a distinct training phase. The resulting network resembles a static entity of knowledge, with endeavours to extend this knowledge without targeting the original task resulting in a catastrophic forgetting. Continual learning shifts this paradigm towards networks that can continually accumulate knowledge over different tasks without the need to retrain from scratch. We focus on task incremental classification, where tasks arrive sequentially and are delineated by clear boundaries. Our main contributions concern 1) a taxonomy and extensive overview of the state-of-the-art, 2) a novel framework to continually determine the stability-plasticity trade-off of the continual learner, 3) a comprehensive experimental comparison of 11 state-of-the-art continual learning methods and 4 baselines. We empirically scrutinize method strengths and weaknesses on three benchmarks, considering Tiny Imagenet and large-scale unbalanced iNaturalist and a sequence of recognition datasets. We study the influence of model capacity, weight decay and dropout regularization, and the order in which the tasks are presented, and qualitatively compare methods in terms of required memory, computation time and storage.

Multi-Task Learning for Dense Prediction Tasks: A Survey.

Abstract: With the advent of deep learning, many dense prediction tasks, i.e. tasks that produce pixel-level predictions, have seen significant performance improvements. The typical approach is to learn these tasks in isolation, that is, a separate neural network is trained for each individual task. Yet, recent multi-task learning (MTL) techniques have shown promising results w.r.t. performance, computations and/or memory footprint, by jointly tackling multiple tasks through a learned shared representation. In this survey, we provide a well-rounded view on state-of-the-art deep learning approaches for MTL in computer vision, explicitly emphasizing on dense prediction tasks. Our contributions concern the following. First, we consider MTL from a network architecture point-of-view. We include an extensive overview and discuss the advantages/disadvantages of recent popular MTL models. Second, we examine various optimization methods to tackle the joint learning of multiple tasks. We summarize the qualitative elements of these works and explore their commonalities and differences. Finally, we provide an extensive experimental evaluation across a variety of dense prediction benchmarks to examine the pros and cons of the different methods, including both architectural and optimization based strategies.

How does Machine Learning Change Software Development Practices

Abstract: Adding an ability for a system to learn inherently adds uncertainty into the system. Given the rising popularity of incorporating machine learning into systems, we wondered how the addition alters software development practices. We performed a mixture of qualitative and quantitative studies with 14 interviewees and 342 survey respondents from 26 countries across four continents to elicit significant differences between the development of machine learning systems and the development of non-machine-learning systems. Our study uncovers significant differences in various aspects of software engineering (e.g., requirements, design, testing, and process) and work characteristics (e.g., skill variety, problem solving and task identity). Based on our findings, we highlight future research directions and provide recommendations for practitioners.

Classification of Hyperspectral Images via Multitask Generative Adversarial Networks

Abstract: Deep learning has shown its huge potential in the field of hyperspectral image (HSI) classification However, most of the deep learning models heavily depend on the quantity of available training samples In this article, we propose a multitask generative adversarial network (MTGAN) to alleviate this issue by taking advantage of the rich information from unlabeled samples Specifically, we design a generator network to simultaneously undertake two tasks: the reconstruction task and the classification task The former task aims at reconstructing an input hyperspectral cube, including the labeled and unlabeled ones, whereas the latter task attempts to recognize the category of the cube Meanwhile, we construct a discriminator network to discriminate the input sample coming from the real distribution or the reconstructed one Through an adversarial learning method, the generator network will produce real-like cubes, thus indirectly improving the discrimination and generalization ability of the classification task More importantly, in order to fully explore the useful information from shallow layers, we adopt skip-layer connections in both reconstruction and classification tasks The proposed MTGAN model is implemented on three standard HSIs, and the experimental results show that it is able to achieve higher performance than other state-of-the-art deep learning models

Studying the Usage of Text-To-Text Transfer Transformer to Support Code-Related Tasks

Abstract: Deep learning (DL) techniques are gaining more and more attention in the software engineering community. They have been used to support several code-related tasks, such as automatic bug fixing and code comments generation. Recent studies in the Natural Language Processing (NLP) field have shown that the Text-To-Text Transfer Transformer (T5) architecture can achieve state-of-the-art performance for a variety of NLP tasks. The basic idea behind T5 is to first pre-train a model on a large and generic dataset using a self-supervised task (e.g., filling masked words in sentences). Once the model is pre-trained, it is fine-tuned on smaller and specialized datasets, each one related to a specific task (e.g., language translation, sentence classification). In this paper, we empirically investigate how the T5 model performs when pre-trained and fine-tuned to support code-related tasks. We pre-train a T5 model on a dataset composed of natural language English text and source code. Then, we fine-tune such a model by reusing datasets used in four previous works that used DL techniques to: (i) fix bugs, (ii) inject code mutants, (iii) generate assert statements, and (iv) generate code comments. We compared the performance of this single model with the results reported in the four original papers proposing DL-based solutions for those four tasks. We show that our T5 model, exploiting additional data for the self-supervised pre-training phase, can achieve performance improvements over the four baselines.

Biobjective Task Scheduling for Distributed Green Data Centers

Abstract: The industry of data centers is the fifth largest energy consumer in the world. Distributed green data centers (DGDCs) consume 300 billion kWh per year to provide different types of heterogeneous services to global users. Users around the world bring revenue to DGDC providers according to actual quality of service (QoS) of their tasks. Their tasks are delivered to DGDCs through multiple Internet service providers (ISPs) with different bandwidth capacities and unit bandwidth price. In addition, prices of power grid, wind, and solar energy in different GDCs vary with their geographical locations. Therefore, it is highly challenging to schedule tasks among DGDCs in a high-profit and high-QoS way. This work designs a multiobjective optimization method for DGDCs to maximize the profit of DGDC providers and minimize the average task loss possibility of all applications by jointly determining the split of tasks among multiple ISPs and task service rates of each GDC. A problem is formulated and solved with a simulated-annealing-based biobjective differential evolution (SBDE) algorithm to obtain an approximate Pareto-optimal set. The method of minimum Manhattan distance is adopted to select a knee solution that specifies the Pareto-optimal task service rates and task split among ISPs for DGDCs in each time slot. Real-life data-based experiments demonstrate that the proposed method achieves lower task loss of all applications and larger profit than several existing scheduling algorithms. Note to Practitioners —This work aims to maximize the profit and minimize the task loss for DGDCs powered by renewable energy and smart grid by jointly determining the split of tasks among multiple ISPs. Existing task scheduling algorithms fail to jointly consider and optimize the profit of DGDC providers and QoS of tasks. Therefore, they fail to intelligently schedule tasks of heterogeneous applications and allocate infrastructure resources within their response time bounds. In this work, a new method that tackles drawbacks of existing algorithms is proposed. It is achieved by adopting the proposed SBDE algorithm that solves a multiobjective optimization problem. Simulation experiments demonstrate that compared with three typical task scheduling approaches, it increases profit and decreases task loss. It can be readily and easily integrated and implemented in real-life industrial DGDCs. The future work needs to investigate the real-time green energy prediction with historical data and further combine prediction and task scheduling together to achieve greener and even net-zero-energy data centers.

The Ordinal Nature of Emotions: An Emerging Approach

Abstract: Computational representation of everyday emotional states is a challenging task and, arguably, one of the most fundamental for affective computing. Standard practice in emotion annotation is to ask people to assign a value of intensity or a class value to each emotional behavior they observe. Psychological theories and evidence from multiple disciplines including neuroscience, economics and artificial intelligence, however, suggest that the task of assigning reference-based values to subjective notions is better aligned with the underlying representations. This paper draws together the theoretical reasons to favor ordinal labels for representing and annotating emotion, reviewing the literature across several disciplines. We go on to discuss good and bad practices of treating ordinal and other forms of annotation data and make the case for preference learning methods as the appropriate approach for treating ordinal labels. We finally discuss the advantages of ordinal annotation with respect to both reliability and validity through a number of case studies in affective computing, and address common objections to the use of ordinal data. More broadly, the thesis that emotions are by nature ordinal is supported by both theoretical arguments and evidence, and opens new horizons for the way emotions are viewed, represented and analyzed computationally.

Learning-Based URLLC-Aware Task Offloading for Internet of Health Things

Abstract: In the Internet of Health Things (IoHT)-based e-Health paradigm, a large number of computational-intensive tasks have to be offloaded from resource-limited IoHT devices to proximal powerful edge servers to reduce latency and improve energy efficiency. However, the lack of global state information (GSI), the adversarial competition among multiple IoHT devices, and the ultra reliable and low latency communication (URLLC) constraints have imposed new challenges for task offloading optimization. In this article, we formulate the task offloading problem as an adversarial multi-armed bandit (MAB) problem. In addition to the average-based performance metrics, bound violation probability, occurrence probability of extreme events, and statistical properties of excess values are employed to characterize URLLC constraints. Then, we propose a URLLC-aware Task Offloading scheme based on the exponential-weight algorithm for exploration and exploitation (EXP3) named UTO-EXP3. URLLC awareness is achieved by dynamically balancing the URLLC constraint deficits and energy consumption through online learning. We provide a rigorous theoretical analysis to show that guaranteed performance with a bounded deviation can be achieved by UTO-EXP3 based on only local information. Finally, the effectiveness and reliability of UTO-EXP3 are validated through simulation results.

Blockchain and Learning-Based Secure and Intelligent Task Offloading for Vehicular Fog Computing

Abstract: Vehicular fog computing has emerged as a complementary framework for edge computing by leveraging the under-utilized computational resources of vehicles. However, how to reduce task offloading delay, queuing delay, and handover cost with incomplete information while simultaneously ensuring privacy, fairness, and security remains an open issue. In this paper, we develop a secure and intelligent task offloading framework to address these challenges. We exploit blockchain and smart contract to facilitate fair task offloading and mitigate various security attacks. Then, we design a subjective logic-based trustfulness metric to quantify the possibility of task offloading success, and develop a trustfulness assessment mechanism. An online learning-based intelligent task offloading algorithm named QUeuing-delay aware, handOver-cost aware, and Trustfulness Aware Upper Confidence Bound (QUOTA-UCB) is proposed, which can learn the long-term optimal strategy and achieve a well-balanced tradeoff among task offloading delay, queuing delay, and handover cost. Finally, extensive theoretical analysis and simulations are carried out to demonstrate the reliability, feasibility, and efficiency of the proposed secure and intelligent task offloading scheme.

Anchor-Free Person Search

Abstract: Person search aims to simultaneously localize and identify a query person from realistic, uncropped images, which can be regarded as the unified task of pedestrian detection and person re-identification (re-id). Most existing works employ two-stage detectors like Faster-RCNN, yielding encouraging accuracy but with high computational overhead. In this work, we present the Feature-Aligned Person Search Network (AlignPS), the first anchor-free framework to efficiently tackle this challenging task. AlignPS explicitly addresses the major challenges, which we summarize as the misalignment issues in different levels (i.e., scale, region, and task), when accommodating an anchor-free detector for this task. More specifically, we propose an aligned feature aggregation module to generate more discriminative and robust feature embeddings by following a "re-id first" principle. Such a simple design directly improves the baseline anchor-free model on CUHK-SYSU by more than 20% in mAP. Moreover, AlignPS outperforms state-of-the-art two-stage methods, with a higher speed. The code is available at https://github.com/daodaofr/AlignPS.

Deep Reinforcement Learning for Delay-Oriented IoT Task Scheduling in SAGIN

Abstract: In this article, we investigate a computing task scheduling problem in space-air-ground integrated network (SAGIN) for delay-oriented Internet of Things (IoT) services. In the considered scenario, an unmanned aerial vehicle (UAV) collects computing tasks from IoT devices and then makes online offloading decisions, in which the tasks can be processed at the UAV or offloaded to the nearby base station or the remote satellite. Our objective is to design a task scheduling policy that minimizes offloading and computing delay of all tasks given the UAV energy capacity constraint. To this end, we first formulate the online scheduling problem as an energy-constrained Markov decision process (MDP). Then, considering the task arrival dynamics, we develop a novel deep risk-sensitive reinforcement learning algorithm. Specifically, the algorithm evaluates the risk, which measures the energy consumption that exceeds the constraint, for each state and searches the optimal parameter weighing the minimization of delay and risk while learning the optimal policy. Extensive simulation results demonstrate that the proposed algorithm can reduce the task processing delay by up to 30% compared to probabilistic configuration methods while satisfying the UAV energy capacity constraint.

Mobility-Aware Joint Task Scheduling and Resource Allocation for Cooperative Mobile Edge Computing

Abstract: Mobile edge computing (MEC) has emerged as a new paradigm to assist low latency services by enabling computation offloading at the network edge. Nevertheless, human mobility can significantly impact the offloading decision and performance in MEC networks. In this context, we propose device-to-device (D2D) cooperation based MEC to expedite the task execution of mobile user by leveraging proximity-aware task offloading. However, user mobility in such distributed architecture results in dynamic offloading decision that instigates mobility-aware task scheduling in our proposed framework. We jointly formulate task assignment and power allocation to minimize the total task execution latency by taking account of user mobility, distributed resources, tasks properties, and energy constraint of the user device. We first propose Genetic Algorithm (GA)-based evolutionary scheme to solve our formulated mixed-integer non-linear programming (MINLP) problem. Then we propose a heuristic named mobility-aware task scheduling (MATS) to obtain effective task assignment with low complexity. The extensive evaluation under realistic human mobility trajectories provides useful insights into the performance of our schemes and demonstrates that, both GA and MATS achieve better latency than other baseline schemes while satisfying the energy constraint of mobile device.

Multi-Task Deep Learning for Real-Time 3D Human Pose Estimation and Action Recognition

Abstract: Human pose estimation and action recognition are related tasks since both problems are strongly dependent on the human body representation and analysis. Nonetheless, most recent methods in the literature handle the two problems separately. In this article, we propose a multi-task framework for jointly estimating 2D or 3D human poses from monocular color images and classifying human actions from video sequences. We show that a single architecture can be used to solve both problems in an efficient way and still achieves state-of-the-art or comparable results at each task while running with a throughput of more than 100 frames per second. The proposed method benefits from high parameters sharing between the two tasks by unifying still images and video clips processing in a single pipeline, allowing the model to be trained with data from different categories simultaneously and in a seamlessly way. Additionally, we provide important insights for end-to-end training the proposed multi-task model by decoupling key prediction parts, which consistently leads to better accuracy on both tasks. The reported results on four datasets (MPII, Human3.6M, Penn Action and NTU RGB+D) demonstrate the effectiveness of our method on the targeted tasks. Our source code and trained weights are publicly available at https://github.com/dluvizon/deephar .

PACE: Privacy-Preserving and Quality-Aware Incentive Mechanism for Mobile Crowdsensing

Abstract: Providing appropriate monetary rewards is an efficient way for mobile crowdsensing to motivate the participation of task participants. However, a monetary incentive mechanism is generally challenging to prevent malicious task participants and a dishonest task requester. Moreover, prior quality-aware incentive schemes are usually failed to preserve the privacy of task participants. Meanwhile, most existing privacy-preserving incentive schemes ignore the data quality of task participants. To tackle these issues, we propose a privacy-preserving and data quality-aware incentive scheme, called PACE. In particular, data quality consists of the reliability and deviation of data. Specifically, we first propose a zero-knowledge model of data reliability estimation that can protect data privacy while assessing data reliability. Then, we quantify the data quality based on the deviation between reliable data and the ground truth. Finally, we distribute monetary rewards to task participants according to their data quality. To demonstrate the effectiveness and efficiency of PACE, we evaluate it in a real-world dataset. The evaluation and analysis results show that PACE can prevent malicious behaviors of task participants and a task requester, and achieves both privacy-preserving and data quality measurement of task participants.

Exploiting Multi-Dimensional Task Diversity in Distributed Auctions for Mobile Crowdsensing

Abstract: To promote development of Mobile Crowdsensing Systems (MCSs), numerous auction schemes have been proposed to motivate mobile users’ participation. But, task diversity of MCSs has not been fully explored by most existing works. To further exploit task diversity and improve performance of MCSs, in this paper, we investigate the joint problem of sensing task assignment and schedule with considering multi-dimensional task diversity, including partial fulfillment, bilaterally-multi-schedule, attribute diversity, and price diversity. First, task owner-centric auction model is formulated and two distributed auction schemes (CPAS and TPAS) are proposed such that each task owner can locally process auction procedure. Then, mobile user-centric auction model is established and two distributed auction schemes (VPAS and DPAS) are developed to facilitate local auction implementation. These four auction schemes differ in their approaches to determine winners and compute payments. We further rigorously prove that all the four auction schemes (CPAS, TPAS, VPAS, and DPAS) are computationally-efficient, individually-rational, and incentive-compatible and that both CPAS and TPAS are budget-feasible. Finally, we comprehensively evaluate the effectiveness of CPAS, TPAS, VPAS, and DPAS via comparing with the state-of-the-art in real-data experiments.

Learning-Based Queue-Aware Task Offloading and Resource Allocation for Space–Air–Ground-Integrated Power IoT

Abstract: Space–air–ground-integrated power Internet of Things (SAG-PIoT) can provide ubiquitous communication and computing services for PIoT devices deployed in remote areas In SAG-PIoT, the tasks can be either processed locally by PIoT devices, offloaded to edge servers through unmanned aerial vehicles (UAVs), or offloaded to cloud servers through satellites However, the joint optimization of task offloading and computational resource allocation faces several challenges, such as incomplete information, dimensionality curse, and coupling between long-term constraints of queuing delay and short-term decision making In this article, we propose a learning-based queue-aware task offloading and resource allocation algorithm (QUARTER) Specifically, the joint optimization problem is decomposed into three deterministic subproblems: 1) device-side task splitting and resource allocation; 2) task offloading; and 3) server-side resource allocation The first subproblem is solved by the Lagrange dual decomposition For the second subproblem, we propose a queue-aware actor–critic-based task offloading algorithm to cope with dimensionality curse A greedy-based low-complexity algorithm is developed to solve the third subproblem Compared with existing algorithms, simulation results demonstrate that QUARTER has superior performances in energy consumption, queuing delay, and convergence

End-to-End Audiovisual Speech Recognition System With Multitask Learning

Abstract: An automatic speech recognition (ASR) system is a key component in current speech-based systems. However, the surrounding acoustic noise can severely degrade the performance of an ASR system. An appealing solution to address this problem is to augment conventional audio-based ASR systems with visual features describing lip activity. This paper proposes a novel end-to-end, multitask learning (MTL), audiovisual ASR (AV-ASR) system. A key novelty of the approach is the use of MTL, where the primary task is AV-ASR, and the secondary task is audiovisual voice activity detection (AV-VAD). We obtain a robust and accurate audiovisual system that generalizes across conditions. By detecting segments with speech activity, the AV-ASR performance improves as its connectionist temporal classification (CTC) loss function can leverage from the AV-VAD alignment information. Furthermore, the end-to-end system learns from the raw audiovisual inputs a discriminative high-level representation for both speech tasks, providing the flexibility to mine information directly from the data. The proposed architecture considers the temporal dynamics within and across modalities, providing an appealing and practical fusion scheme. We evaluate the proposed approach on a large audiovisual corpus (over 60 hours), which contains different channel and environmental conditions, comparing the results with competitive single task learning (STL) and MTL baselines. Although our main goal is to improve the performance of our ASR task, the experimental results show that the proposed approach can achieve the best performance across all conditions for both speech tasks. In addition to state-of-the-art performance in AV-ASR, the proposed solution can also provide valuable information about speech activity, solving two of the most important tasks in speech-based applications.

Pareto Self-Supervised Training for Few-Shot Learning

Abstract: While few-shot learning (FSL) aims for rapid generalization to new concepts with little supervision, self-supervised learning (SSL) constructs supervisory signals directly computed from unlabeled data. Exploiting the complementarity of these two manners, few-shot auxiliary learning has recently drawn much attention to deal with few labeled data. Previous works benefit from sharing inductive bias between the main task (FSL) and auxiliary tasks (SSL), where the shared parameters of tasks are optimized by minimizing a linear combination of task losses. However, it is challenging to select a proper weight to balance tasks and reduce task conflict. To handle the problem as a whole, we propose a novel approach named as Pareto self-supervised training (PSST) for FSL. PSST explicitly decomposes the few-shot auxiliary problem into multiple constrained multi-objective subproblems with different trade-off preferences, and here a preference region in which the main task achieves the best performance is identified. Then, an effective preferred Pareto exploration is proposed to find a set of optimal solutions in such a preference region. Extensive experiments on several public benchmark datasets validate the effectiveness of our approach by achieving state-of-the-art performance.

Optimizing Task Assignment for Reliable Blockchain-Empowered Federated Edge Learning

Abstract: A rapid-growing machine learning technique called federated edge learning has emerged to allow a massive number of edge devices (eg smart phones) to collaboratively train globally shared models without revealing their private raw data This technique not only ensures good machine learning performance but also maintains data privacy of the edge devices However, the federated edge learning still faces the following critical challenges: (i) difficulty in avoiding unreliable edge devices acting as workers for federated edge learning, and (ii) lack of efficient learning task assignment schemes among task publishers and workers To tackle these challenges, reputation is utilized as a metric to evaluate the trustworthiness and reliability of the edge devices A many-to-one matching model is proposed to address the task assignment problem between task publishers and reliable workers with high reputation For stimulating reliable edge devices to join model training and enable secure reputation management, blockchain is employed to store the training records and manage reputation data in a decentralized and secure manner without the risk of a single point of failure Numerical results show that the proposed schemes can achieve significant performance improvement in terms of reliability of federated edge learning

Matching-Theory-Based Low-Latency Scheme for Multitask Federated Learning in MEC Networks

Abstract: Nowadays, there is an ever-increasing interests in federated learning, which allows end devices to collaboratively train a global machine learning model in a decentralized paradigm without sharing individual data. Despite the advantages of low communication cost and preserving data privacy, federated learning is also facing with new challenges to address. Practically, end devices will consider the resources cost and willingness caused by machine learning model training when they are invited to participate a federated learning task. So, how to assign the preferable tasks to the devices with high willingness has to be considered. Besides, the end devices have the property of high mobility, which means the time of devices localizing within the network is limited. Therefore, to reduce the task execution time is necessary. To address these problems, we first analyze and formulate the latency minimization problem for multitask federated learning in a multiaccess edge computing (MEC) network scenario. Then, we model the corresponding problem as a matching game to find the optimal task assignment solutions. Moreover, considering the large-scale Internet-of-Things (IoT) scenario, it is almost impossible for two sides to know the details of every individual of the other side so that the complete preference list (CPL) cannot be built in reality. Therefore, we propose an algorithm for large-scale matching with the incomplete preference list to address the problem. Finally, we conduct the numerical simulation in various cases to demonstrate the effectiveness of our proposed method. The results show that our approach can achieve similar performance with the CPL case.

Multi-Task Allocation Under Time Constraints in Mobile Crowdsensing

Abstract: Mobile crowdsensing (MCS) is a popular paradigm to collect sensed data for numerous sensing applications. With the increment of tasks and workers in MCS, it has become indispensable to design efficient task allocation schemes to achieve high performance for MCS applications. Many existing works on task allocation focus on single-task allocation, which is inefficient in many MCS scenarios where workers are able to undertake multiple tasks. On the other hand, many tasks are time-limited, while the available time of workers is also limited. Therefore, time validity is essential for both tasks and workers. To accommodate these challenges, this paper proposes a multi-task allocation problem with time constraints, which investigates the impact of time constraints to multi-task allocation and aims to maximize the utility of the MCS platform. We first prove that this problem is NP-complete. Then two evolutionary algorithms are designed to solve this problem. Finally, we conduct the experiments based on synthetic and real-world datasets under different experiment settings. The results verify that the proposed algorithms achieve more competitive and stable performance compared with baseline algorithms.

Energy-Efficient Task Offloading and Resource Allocation via Deep Reinforcement Learning for Augmented Reality in Mobile Edge Networks

Abstract: The augmented reality (AR) applications have been widely used in the field of Internet of Things (IoT) because of good immersion experience for users, but their ultralow delay demand and high energy consumption bring a huge challenge to the current communication system and terminal power. The emergence of mobile-edge computing (MEC) provides a good thinking to solve this challenge. In this article, we study an energy-efficient task offloading and resource allocation scheme for AR in both the single-MEC and multi-MEC systems. First, a more specific and detailed AR application model is established as a directed acyclic graph according to its internal functionality. Second, based on this AR model, a joint optimization problem of task offloading and resource allocation is formulated to minimize the energy consumption of each user subject to the latency requirement and the limited resources. The problem is a mixed multiuser competition and cooperation problem, which involves the task offloading decision, uplink/downlink transmission resources allocation, and computing resources allocation of users and MEC server. Since it is an NP-hard problem and the communication environment is dynamic, it is difficult for genetic algorithms or heuristic algorithms to solve. Therefore, we propose an intelligent and efficient resource allocation and task offloading algorithm based on the deep reinforcement learning framework of multiagent deep deterministic policy gradient (MADDPG) in a dynamic communication environment. Finally, simulation results show that the proposed algorithm can greatly reduce the energy consumption of each user terminal.

Integrated Task Assignment and Path Planning for Capacitated Multi-Agent Pickup and Delivery

Abstract: Multi-agent Pickup and Delivery (MAPD) is a challenging industrial problem where a team of robots is tasked with transporting a set of tasks, each from an initial location and each to a specified target location. Appearing in the context of automated warehouse logistics and automated mail sortation, MAPD requires first deciding which robot is assigned what task (i.e., Task Assignment or TA) followed by a subsequent coordination problem where each robot must be assigned collision-free paths so as to successfully complete its assignment (i.e., Multi-Agent Path Finding or MAPF). Leading methods in this area solve MAPD sequentially: first assigning tasks, then assigning paths. In this work we propose a new coupled method where task assignment choices are informed by actual delivery costs instead of by lower-bound estimates. The main ingredients of our approach are a marginal-cost assignment heuristic and a meta-heuristic improvement strategy based on Large Neighbourhood Search. As a further contribution, we also consider a variant of the MAPD problem where each robot can carry multiple tasks instead of just one. Numerical simulations show that our approach yields efficient and timely solutions and we report significant improvement compared with other recent methods from the literature.

Distributed Task Migration Optimization in MEC by Extending Multi-Agent Deep Reinforcement Learning Approach

Abstract: Closer to mobile users geographically, mobile edge computing (MEC) can provide some cloud-like capabilities to users more efficiently. This enables it possible for resource-limited mobile users to offload their computation-intensive and latency-sensitive tasks to MEC nodes. For its great benefits, MEC has drawn wide attention and extensive works have been done. However, few of them address task migration problem caused by distributed user mobility, which can’t be ignored with quality of service (QoS) consideration. In this article, we study task migration problem and try to minimize the average completion time of tasks under migration energy budget. There are multiple independent users and the movement of each mobile user is memoryless with a sequential decision-making process, thus reinforcement learning algorithm based on Markov chain model is applied with low computation complexity. To further facilitate cooperation among users, we devise a distributed task migration algorithm based on counterfactual multi-agent (COMA) reinforcement learning approach to solve this problem. Extensive experiments are carried out to assess the performance of this distributed task migration algorithm. Compared with no migrating (NM) and single-agent actor-critic (AC) algorithms, the proposed distributed task migration algorithm can achieve up 30-50 percent reduction about average completion time.

Privacy-Enhanced Data Fusion for COVID-19 Applications in Intelligent Internet of Medical Things

Abstract: With the worldwide large-scale outbreak of COVID-19, the Internet of Medical Things (IoMT), as a new type of Internet of Things (IoT)-based intelligent medical system, is being used for COVID-19 prevention and detection. However, since the widespread use of IoMT will generate a large amount of sensitive information related to patients, it is becoming more and more important yet challenging to ensure data security and privacy of COVID-19 applications in IoMT. The leakage of private information during IoMT data fusion process will cause serious problems and affect people’s willingness to contribute data in IoMT. To address these challenges, this article proposes a new privacy-enhanced data fusion strategy (PDFS). The proposed PDFS consists of four important components, i.e., sensitive task classification, task completion assessment, incentive mechanism-based task contract design, and homomorphic encryption-based data fusion. The extensive simulation experiments demonstrate that PDFS can achieve high task classification accuracy, task completion rate, task data reliability and task participation rate, and low average error rate, while improving the privacy protection for data fusion under COVID-19 application environments based on IoMT.

Recovery RL: Safe Reinforcement Learning With Learned Recovery Zones

Abstract: Safety remains a central obstacle preventing widespread use of RL in the real world: learning new tasks in uncertain environments requires extensive exploration, but safety requires limiting exploration. We propose Recovery RL, an algorithm which navigates this tradeoff by (1) leveraging offline data to learn about constraint violating zones before policy learning and (2) separating the goals of improving task performance and constraint satisfaction across two policies: a task policy that only optimizes the task reward and a recovery policy that guides the agent to safety when constraint violation is likely. We evaluate Recovery RL on 6 simulation domains, including two contact-rich manipulation tasks and an image-based navigation task, and an image-based obstacle avoidance task on a physical robot. We compare Recovery RL to 5 prior safe RL methods which jointly optimize for task performance and safety via constrained optimization or reward shaping and find that Recovery RL outperforms the next best prior method across all domains. Results suggest that Recovery RL trades off constraint violations and task successes 2–20 times more efficiently in simulation domains and 3 times more efficiently in physical experiments. See https://tinyurl.com/rl-recovery for videos and supplementary material.

Density-Aware Multi-Task Learning for Crowd Counting

Abstract: In this paper, we present a method called density-aware convolutional neural network (DensityCNN) to perform the crowd counting task in various crowded scenes. The key idea of the DensityCNN is to utilize high-level semantic information to provide guidance and constraint when generating density maps. To this end, we implement the DensityCNN by adopting a multi-task CNN structure to jointly learn density-level classification and density map estimation. The density-level classification task learns multi-channel semantic features that are aware of the density distributions of the input image. This task is accomplished via our specially designed group-based convolutional structure in a supervised learning manner. In the density map estimation task, these semantic features are deployed together with high-dimension convolutional features to generate density maps with lower count errors. Extensive experiments on four challenging crowd datasets (ShanghaiTech, UCF_CC_50, UCF-QNCF, and WorldExpo’10) and one vehicle dataset TRANCOS demonstrate the effectiveness of the proposed method.

Joint Task Scheduling and Containerizing for Efficient Edge Computing

Abstract: Container-based operation system (OS) level virtualization has been adopted by many edge-computing platforms. However, for an edge server, inter-container communications, and container management consume significant CPU resources. Given an application composed of interdependent tasks, the number of such operations is closely related to the dependency between the scheduled tasks. Thus, to improve the execution efficiency of an application in an edge server, task scheduling and task containerizing need to be considered together. To this end, a joint task scheduling and containerizing (JTSC) scheme is developed in this article. Experiments are first carried out to quantify the resource utilization of container operations. System models are then built to capture the features of task execution in containers in an edge server with multiple processors. With these models, joint task scheduling and containerizing is conducted as follows. First, tasks are scheduled without considering containerization, which results in initial schedules. Second, based on system models and guidelines gained from the initial schedules, several containerization algorithms are designed to map tasks to containers. Third, task execution durations are updated by adding the time for inter-container communications, and then the task schedules are updated accordingly. The JTSC scheme is evaluated through extensive simulations. The results show that it reduces inefficient container operations and enhances the execution efficiency of applications by 60 percent.

A Machine Learning Approach for Task and Resource Allocation in Mobile-Edge Computing-Based Networks

Abstract: In this article, a joint task, spectrum, and transmit power allocation problem is investigated for a wireless network in which the base stations (BSs) are equipped with mobile-edge computing (MEC) servers to jointly provide computational and communication services to users. Each user can request one computational task from three types of computational tasks. Since the data size of each computational task is different, as the requested computational task varies, the BSs must adjust their resource (subcarrier and transmit power) and task allocation schemes to effectively serve the users. This problem is formulated as an optimization problem whose goal is to minimize the maximal computational and transmission delay among all users. A multistack reinforcement learning (RL) algorithm is developed to solve this problem. Using the proposed algorithm, each BS can record the historical resource allocation schemes and users’ information in its multiple stacks to avoid learning the same resource allocation scheme and users’ states, thus improving the convergence speed and learning efficiency. The simulation results illustrate that the proposed algorithm can reduce the number of iterations needed for convergence and the maximal delay among all users by up to 18% and 11.1% compared to the standard $Q$ -learning algorithm.

Parking Edge Computing: Parked-Vehicle-Assisted Task Offloading for Urban VANETs

Abstract: Vehicular edge computing has been a promising paradigm to offer low-latency and high reliability vehicular services for users. Nevertheless, for compute-intensive vehicle applications, most previous researches cannot perform them efficiently due to both the inadequate of infrastructure construction and the computing resource bottleneck of the edge server. Motivated by the fact that there is a large number of outside parked vehicles with rich and underutilized resources in the urban area, we propose the idea of parking edge computing, which makes use of the parked vehicles to assist edge servers in offloaded task handling. Specifically, on-street and off-street parked vehicles are first organized into parking clusters to act as virtual edge servers, participating in offloaded tasks execution in our framework. Second, a novel task scheduling algorithm is designed to jointly decide edge server selection and resource assignment. Furthermore, a local task scheduling policy is proposed as well, which reasonably allocates parked vehicles to perform the tasks with the aim of further improving task offloading performance. Finally, a time-related trajectory prediction model based on the random forest model is built, which helps to send back output result accurately. Our framework not only requires no additional infrastructure investment but also provides adequate computing resources. Simulation results based on a real city map and realistic traffic situations demonstrate that our framework provides more efficient and stable offloading services, especially in a large number of task requests condition.