The advancement of intelligent agents has revolutionized problem-solving across diverse domains, yet solutions for personalized fashion styling remain underexplored, which holds immense promise for promoting shopping experiences. In this work, we present StyleTailor, the first collaborative agent framework that seamlessly unifies personalized apparel design, shopping recommendation, virtual try-on, and systematic evaluation into a cohesive workflow. To this end, StyleTailor pioneers an iterative visual refinement paradigm driven by multi-level negative feedback, enabling adaptive and precise user alignment. Specifically, our framework features two core agents, i.e., Designer for personalized garment selection and Consultant for virtual try-on, whose outputs are progressively refined via hierarchical vision-language model feedback spanning individual items, complete outfits, and try-on efficacy. Counterexamples are aggregated into negative prompts, forming a closed-loop mechanism that enhances recommendation quality. To assess the performance, we introduce a comprehensive evaluation suite encompassing style consistency, visual quality, face similarity, and artistic appraisal. Extensive experiments demonstrate StyleTailor's superior performance in delivering personalized designs and recommendations, outperforming strong baselines without negative feedback and establishing a new benchmark for intelligent fashion systems.

The rapid adoption of electric vehicles (EVs) has created new opportunities for decentralized energy trading, where EVs can act as mobile energy providers in peer-to-peer markets. Blockchain provides a secure foundation for such systems, ensuring trust and accountability. However, its inherent transparency creates privacy risks, as it enables the tracking of trading activities. Existing privacy-preserving mechanisms typically focus on concealing payment transactions but often expose other critical interactions, such as matching coordination. To address these challenges, we proposePriVET, a privacy-preserving framework for vehicular energy trading. PriVET leverages smart contracts for trade matching and uses an enhanced Paillier encryption scheme to support encrypted comparisons, ensuring secure coordination without revealing sensitive data. Additionally, a Bloom-filter– based Geohash encoding is used to protect location privacy during spatial matching. We evaluate PriVET through both theoretical analysis and practical experiments. In a simulation environment, the transaction computation time for 100 vehicles is shown to be under 30ms, with communication overhead kept below 20KB. These results demonstrate that PriVET provides robust privacy protection while maintaining minimal overhead, making it a practical solution for real-world blockchain-based energy trading scenarios.

Efficient and reliable data collection is essential for agricultural Internet of Things (IoT) systems, where timely sensing supports precision farming. Unmanned aerial vehicle (UAV)-assisted collection provides flexible and low cost coverage. However, in wide and remote fields, UAV-assisted collection faces two coupled challenges. First, agricultural field operations and field-state changes, such as irrigation, rainfall, harvesting, canopy occlusion, machinery movement, and fluctuations in solar exposure, can jointly and unevenly perturb traffic, communication, and energy processes, leading to bursty arrivals, degraded air-to-ground (A2G) links, and variations in harvested energy. Second, under single-link communication, the UAV can directly access only one node in each slot, making frequent global state refresh difficult and causing stale beliefs to accumulate rapidly after abrupt changes. These effects lead to biased scheduling, unnecessary maneuvering, packet loss, and excess energy consumption. To address this problem, this paper proposes a three stage framework for joint node scheduling and UAV trajectory control. It first refreshes selected node states to correct stale beliefs and then encodes refreshed and inferred states through a reliability-aware Transformer. Finally, it performs conservative hybrid-action control to coordinate discrete scheduling and continuous motion while mitigating value overestimation under partial observability and non-stationarity. Simulations under point-shift, cluster-shift, and global-shift scenarios show that the proposed method consistently reduces packet loss, improves energy efficiency, and achieves faster recovery than representative reinforcement learning (RL) and heuristic baselines.

Mobile edge computing (MEC), integrated with directed acyclic graph (DAG) blockchain technology, has emerged as a promising paradigm for ensuring secure and efficient resource trading between IoT user equipment (UEs) and edge service providers (ESPs). However, due to the open and heterogeneous nature of MEC networks, ESPs are susceptible to malicious attacks, rendering resource trading information potentially unreliable. While DAG blockchains ensure the reliability of on-chain data, they fail to guarantee the trustworthiness of ESPs and cannot effectively incentivize their participation in resource trading and blockchain consensus. To address these challenges, we develop a trust-driven resource trading framework for DAG blockchain-aided MEC networks. In this framework, we first design an off-chain trust-driven resource pricing mechanism, in which the resource price set by each ESP is positively correlated with its trust value. In addition, we design an on-chain trust-driven consensus mechanism, wherein the on-chain security of transactions published by each ESP is positively associated with its trust level. To enable UEs to better evaluate the on-chain transaction security, we design a novel metric termed transaction security satisfaction and incorporate it into the utility function of UEs. Furthermore, we model the resource trading between UEs and ESPs as a multi-leader multi-follower Stackelberg game, and verify the existence and uniqueness of its equilibrium. To maximize the utilities of both ESPs and UEs, we propose a backward induction-based iterative algorithm to jointly optimize resource pricing, resource demand, and offloading strategy. Numerical simulations validate the effectiveness of our proposed scheme, demonstrating its superior performance compared with baseline schemes.

The need for enhanced transaction privacy in decentralized finance (DeFi) is critical. However, existing coin mixing solutions often reveal telltale patterns on the blockchain, exposing users to heuristic analysis. This paper presents DeFiMix, an indistinguishable coin mixing scheme engineered to obscure transaction flows while guaranteeing fairness and security. DeFiMix achieves this through a dual-layer mechanism. First, an off-chain secret handshake protocol enables anonymous negotiation between senders and mixers, effectively breaking the link between transactions and participants. Second, on-chain transactions are structured using time-locks and concurrent signatures to resemble common DeFi activities such as staking and lending, rendering them indistinguishable from ordinary operations. Using security analysis and extensive simulations, we validate DeFiMix’s ability to prevent transaction linkage while remaining practically viable. The results underscore DeFiMix’s strong indistinguishability and fairness, alongside its minimal computational demands, establishing it as a compelling solution for privacy-focused transactions within the DeFi ecosystem.

Vehicular twin networks (VTN) create virtual agents of vehicular entities through digital twin (DT) technology, replacing physical counterparts in connecting and exchanging traffic information in cyberspace, overcoming physical range constraints and extending information sources for enhanced vehicular decision support. However, the inherent openness of VTN renders communication between DTs, vulnerable to security threats, such as tampering and impersonation, especially in scenarios where DTs are distributed across multiple cloud domains. These issues result in erroneous decisions to threaten vehicular safety because DTs may receive compromised information. To address these challenges, this article proposes a secure and revocable access authentication scheme in the cross-domain VTN. In the scheme, DTs should be authorized first to obtain identity-bound symmetric functions before joining the VTN, and then perform secure access authentication and key agreement with others based on chameleon hash functions for both intradomain and cross-domain communication. Moreover, a dynamic revocation mechanism is introduced to remove malicious DTs from VTN. Formal verification using the Tamarin tool demonstrates that the proposed scheme achieves diverse security properties. Performance evaluation further shows that the proposed scheme outperforms most related schemes in terms of computational and communication overhead.

Sparse vector coding (SVC) is a promising short-packet transmission method for ultra reliable low latency communication (URLLC) in next generation communication systems. In this paper, a dual-mapping SVC (DM-SVC) based short packet transmission scheme is proposed to further enhance the transmission performance of SVC. The core idea behind the proposed scheme lies in mapping the transmitted information bits onto sparse vectors via block and single-element sparse mappings. The block sparse mapping pattern is able to concentrate the transmit power in a small number of non-zero blocks thus improving the decoding accuracy, while the single-element sparse mapping pattern ensures that the code length does not increase dramatically with the number of transmitted information bits. At the receiver, a two-stage decoding algorithm is proposed to sequentially identify non-zero block indexes and single-element non-zero indexes. Extensive simulation results verify that proposed DM-SVC scheme outperforms the existing SVC schemes in terms of block error rate and spectral efficiency.

Vehicular digital twins (DTs) create virtual representations of physical vehicles, enabling real-time data exchange to enhance intelligence and ensure safe driving. Reducing DT content delivery latency in infrastructure-deficient, sparsely populated areas is crucial. This paper develops a novel Satellite-UAV multi-path content delivery framework for data synchronization in vehicular DT applications. Satellites offer wide coverage but suffer from high latency, while UAVs provide rapid deployment and low-latency communication. The framework leverages these unique characteristics to facilitate simultaneous content downloading through multiple paths, thereby reducing latency. A Stackelberg game model is used to motivate effective resource allocation by UAVs. Given the typically private utility model of DTs, a learning-based algorithm is developed to determine optimal pricing strategies for UAVs. Simulation results demonstrate significant enhancements in UAV utility and reduced DT costs, meeting diverse service requirements.

Vehicular edge computing (VEC) is an emerging computing paradigm that alleviates the limitations of local computing resources for the Internet of Vehicles. However, the lack of trust among distributed nodes and ineffective incentive mechanisms discourage Roadside Units (RSUs) from providing resources. In addition, information asymmetry can lead to the clearing prices of resources failing to accurately reflect the actual value of resources. To address these issues, we propose a trustbased double auction framework in VEC networks. In this framework, to incentivize the RSUs with high trust to participate in resource trading and ensure the clearing prices accurately reflect the actual value of resources, we first design a trustbased resource pricing mechanism. In this mechanism, RSUs with higher trust can set higher resource prices and the clearing prices of resources are closer to the buyers’ bids. Then, we develop a trust-based double auction mechanism that incorporates the Edmonds-Karp algorithm for efficient buyer-seller matching. Considering the time-varying nature of the VEC networks, we employ deep reinforcement learning to optimize decision-making to maximize the social welfare. Finally, we demonstrate that the proposed double auction model satisfies key economic properties such as individual rationality and incentive compatibility. Simulation results validate that our proposed approach outperforms benchmark schemes in terms of social welfare.

Recent advancements in Generalizable Gaussian Splatting have enabled robust 3D reconstruction from sparse input views by utilizing feed-forward Gaussian Splatting models, achieving superior cross-scene generalization. However, while many methods focus on geometric consistency, they often neglect the potential of text-driven guidance to enhance semantic understanding, which is crucial for accurately reconstructing fine-grained details in complex scenes. To address this limitation, we propose TextSplat-the first text-driven Generalizable Gaussian Splatting framework. Specifically, our framework employs three parallel modules to obtain complementary representations: the Diffusion Prior Depth Estimator for accurate depth information, the Semantic Aware Segmentation Network for detailed semantic information, and the Multi-View Interaction Network for refined cross-view features. Then, in the Text-Guided Semantic Fusion Module, these representations are integrated via the text-guided and attention-based feature aggregation mechanism, resulting in enhanced 3D Gaussian parameters enriched with detailed semantic cues. Experimental results on various benchmark datasets demonstrate improved performance compared to existing methods across multiple evaluation metrics, validating the effectiveness of our framework. The code will be publicly available.

Blockchain is considered as a promising technology to ensure the security of resource trading between the IoT user equipment (UEs) and the edge service providers (ESPs) in mobile edge computing (MEC) networks. However, blockchain cannot guarantee the trustworthiness of ESPs and cannot effectively incentivize ESPs to participate in resource trading and blockchain consensus. In addition, the high computational resource demands, energy consumption, and the limited transaction throughput of traditional blockchain pose challenges to IoT applications that require frequent micro transactions. To address these issues, we develop an integrated blockchain and MEC framework based on a directed acyclic graph (DAG) ledger to meet the demands of IoT applications. In this framework, we first model the resource trading between the UEs and the ESP as a multi-follower Stackelberg game. To incentivize the ESP to participate in resource trading and blockchain consensus, we design a trust based resource pricing mechanism, wherein the trust of ESP is evaluated by UEs and the ESP with higher trust can set a higher resource price. Additionally, to enable UEs to better assess the security of transactions on the DAG blockchain, we design a metric called transaction security satisfaction and adopt it as the revenue of UEs. Second, we verify the existence and uniqueness of the Stackelberg equilibrium. Furthermore, we propose a backward induction based iterative algorithm to optimize the resource pricing strategy for ESP and the resource demand strategy for UEs, while maximizing the utilities of both ESP and UEs. Numerical simulations demonstrate the effectiveness of our proposed scheme, showing its superiority over benchmark scheme in terms of transaction security satisfaction and the utilities of ESP and UEs.

Sparse vector coding (SVC) is a promising short-packet transmission method for ultra reliable low latency communication (URLLC) in next generation mobile communication systems. However, the storage burden of codebook and high decoding complexity limit its application in Internet of Things (loT) devices with constrained storage space and computational capabilities. To tackle this challenge, a semi-tensor SVC (ST-SVC)-based short-packet transmission scheme is proposed in this paper. The core idea behind ST-SVC is that it utilizes the semi-tensor product (STP) model in random spreading process, replacing the matrix multiplication model used in traditional SVC schemes. At the transmitter, a low-dimensional codebook is utilized to perform random spreading on a high-dimensional sparse vector carrying information bits. At the receiver, by exploiting the Kronecker structure induced by the STP model, a low-complexity parallel support identification algorithm is proposed for ST-SVC decoding. The proposed scheme breaks through the dimension matching condition required between the codebook matrix and high-dimensional sparse vector in traditional SVC schemes, allowing the loT devices to store an ultra-low-dimensional codebook, which significantly reduces storage overhead. Simulation results demonstrate that the proposed ST-SVC scheme can achieve a substantial reduction in both storage overhead and decoding latency compared to state-of-the-art SVC schemes, with only a slight performance loss in block error rate.

Despite the rapid growth of electric vehicles (EVs), charging remains a time-consuming issue that requires effective management. An important solution uses digital twin (DT) technology, which acts as a virtual agent for EVs in the digital space. DT can analyze real-time vehicle data to develop optimal charging schedules and reserve charging providers in advance through the vehicular network, leading to more efficient charging processes. However, the vehicular network exposes the automated reservation process of the DT to security attacks. Additionally, there is a risk that the actual charging process may deviate from the scheduled requirements set by the DT, resulting in wasted charging resources. To address these issues, this article proposes a secure and efficient charging reservation scheme based on DT technology. To prevent malicious attacks, we first design a secure and privacy-preserving reservation authentication protocol using the extended Chebyshev chaotic maps, taking into account the computational resources of the EV. Furthermore, we develop a reputation mechanism to evaluate and incentivize the charging behavior of EVs. Formal verification and further discussions are conducted to show diverse security functionalities of the proposed scheme can be achieved. We evaluate that the proposed scheme outperforms existing schemes in terms of computation and communication overheads, while also assessing the impact of EV charging behavior on reputation and charging level.

Low Earth Orbit (LEO) satellite networks offer a promising approach for achieving global coverage and high-speed Internet access. However, the existing communication frameworks within these networks may result in low robustness and reliability. This paper introduces QTER, a novel and lightweight routing framework specifically designed for LEO networks, aimed at addressing these challenges. Our contributions are threefold. First, QTER establishes a cost-effective and reliable routing framework that accommodates a variety of applications with distinct Quality of Service (QoS) requirements. By enabling multi-path routing, the framework minimizes costs while ensuring end-to-end reliability, thus enhancing adaptability to diverse service demands. Second, we leverage the unique structural characteristics of LEO satellite networks by modeling the satellite constellation as a three-dimensional network, wherein higher-shell satellites serve as management satellite nodes (MSNs) to coordinate routing strategies among lower-shell satellites. This architecture significantly improves system efficiency and adaptability. Third, we propose a failure recovery mechanism that allows MSNs to relay packets when lower-orbit satellites are rendered unavailable due to environmental factors, thereby enhancing system robustness. Extensive simulations demonstrate that QTER exhibits resilience to node failures and dynamic network conditions, achieving reductions in average cost and delay by 59.4% and 38.9%, respectively, compared to baselines.

With the rapid development and extensive deployment of the fifth-generation wireless system (5G), it has achieved ubiquitous high-speed connectivity and improved overall communication performance. Additionally, as one of the promising technologies for integration beyond 5G, digital twin in cyberspace can interact with the core network, transmit essential information, and further enhance the wireless communication quality of the corresponding mobile device (MD). However, the utilization of millimeter-wave, terahertz band, and ultra-dense network technologies presents urgent challenges for MD in 5G and beyond, particularly in terms of frequent handover authentication with target base stations during faster mobility, which can cause connection interruption and incur malicious attacks. To address such challenges in 5G and beyond, in this paper, we propose a secure and efficient handover authentication scheme by utilizing digital twin. Acting as an intelligent intermediate, the authorized digital twin can handle computations and assist the corresponding MD in performing secure mutual authentication and key negotiation in advance before attaching the target base stations in both intra-domain and inter-domain scenarios. In addition, we provide the formal verification based on BAN logic, RoR model, and ProVerif, and informal analysis to demonstrate that the proposed scheme can offer diverse security functionality. Performance evaluation shows that the proposed scheme outperforms most related schemes in terms of signaling, computation, and communication overheads.

Gait recognition aims to identify individuals based on their body shape and walking patterns. Though much progress has been achieved driven by deep learning, gait recognition in real-world surveillance scenarios remains quite challenging to current methods. Conventional approaches, which rely on periodic gait cycles and controlled environments, struggle with the non-periodic and occluded silhouette sequences encountered in the wild. In this paper, we propose a novel framework,TrackletGait, designed to address these challenges in the wild. We propose Random Tracklet Sampling, a generalization of existing sampling methods, which strikes a balance between robustness and representation in capturing diverse walking patterns. Next, we introduce Haar Wavelet-based Downsampling to preserve information during spatial downsampling. Finally, we present a Hardness Exclusion Triplet Loss, designed to exclude low-quality silhouettes by discarding hard triplet samples. TrackletGait achieves state-of-the-art results, with 77.8% and 80.4% rank-1 accuracy on the Gait3D and GREW datasets, respectively, while using only 10.3M backbone parameters. Extensive experiments are also conducted to further investigate the factors affecting gait recognition in the wild.

The rapid development of Low Earth Orbit (LEO) satellite networks has provided ubiquitous Internet access to users around the world, especially in areas where there are no terrestrial networks. However, a dish can only communicate with one of the available satellites when uploading data in the current framework, resulting in low communication efficiency. As the number of satellites continues to increase, the current framework cannot make full use of the user-satellite link resources. In this paper, we first conduct a measurement of Starlink’s network performance and report some unique features. Then, we propose an adaptive multi-link data allocation framework for LEO satellite networks where a dish can communicate with multiple satellites at the same time to improve data transmission efficiency. With this framework, data can be split into chunks and uploaded simultaneously over multiple links. Our goal is to determine the data allocation strategies to jointly optimize the transmission latency and data processing costs. To this end, we propose a deep reinforcement learning-based algorithm integrated with the traffic prediction module to determine the optimal data allocation strategies in a dynamic network environment. Through extensive simulations, we demonstrate the effectiveness of our approach compared with baselines.

Existing studies for gait recognition primarily utilized sequences of either binary silhouette or human parsing to encode the shapes and dynamics of persons during walking. Silhouettes exhibit accurate segmentation quality and robustness to environmental variations, but their low information entropy may result in sub-optimal performance. In contrast, human parsing provides fine-grained part segmentation with higher information entropy, but the segmentation quality may deteriorate due to the complex environments. To discover the advantages of silhouette and parsing and overcome their limitations, this paper proposes a novel cross-granularity alignment gait recognition method, named XGait, to unleash the power of gait representations of different granularity. To achieve this goal, the XGait first contains two branches of backbone encoders to map the silhouette sequences and the parsing sequences into two latent spaces, respectively. Moreover, to explore the complementary knowledge across the features of two representations, we design the Global Cross-granularity Module (GCM) and the Part Cross-granularity Module (PCM) after the two encoders. In particular, the GCM aims to enhance the quality of parsing features by leveraging global features from silhouettes, while the PCM aligns the dynamics of human parts between silhouette and parsing features using the high information entropy in parsing sequences. In addition, to effectively guide the alignment of two representations with different granularity at the part level, an elaborate-designed learnable division mechanism is proposed for the parsing features. Finally, comprehensive experiments on two large-scale gait datasets not only show the superior performance of XGait with the Rank-1 accuracy of 80.5% on Gait3D and 88.3% CCPG but also reflect the robustness of the learned features even under challenging conditions like occlusions and cloth changes

The evolving digital twin technology translates physical entities into the digital realm, allowing the exploration of abundant digital resources to optimize the task execution of these physical entities. Real-time data synchronization between physical entities and their digital twins is essential for the effective functioning of digital twin systems. In this paper, we investigate the challenge of data synchronization in vehicular digital twin networks operating in open street scenarios, where multiple vehicles rely on cellular networks for continuous data synchronization with their digital twins. Given the contention for cellular bandwidth among vehicles, a coordination scheme is required to manage resource allocation. As vehicles are fully distributed driven by self-interests only, a game-theoretic approach is proposed that leverages a cloud center controller to guide the sharing of cellular resources among digital twins. An optimal incentive mechanism is introduced to encourage digital twins to adhere to the center's guidance, promoting global social welfare. Through extensive simulations, we demonstrate that the proposed scheme successfully motivates vehicles to follow the center's guidance, leading to efficient data synchronization and mutual benefit maximization.

The increasing demand for online high-quality video streaming has brought huge challenges to the traditional client-server video streaming systems due to the high feedback delay, rigorous bandwidth requirement, and the lack of a mechanism of centralized resource management between users. In this work, we propose AMIS-MU, an edge computing-based mobile video streaming system that optimizes the watching experience of users via playback adaptation and channel resource allocation. AMIS-MU fully explores the power of edge servers from three perspectives. First, by pre-caching videos from the cloud, AMIS-MU analyzes video contents at the edge, and achieves a nearly imperceptible content-based playback speed adaptation. Second, as the edge server controls the channel resources of users in a centralized fashion, AMIS-MU adaptively updates the channel configuration to optimize the overall watching experience. Last, the plenty of computational power available at the edge enables a more intelligent playback control by using deep reinforcement learning (DRL). We propose a novel usage of DRL which significantly reduces the complexity of the cross-layer joint optimization problem and solve the non-convex channel resource allocation problem by Lyapunov optimization. Experiments show that AMIS-MU outperforms other existing algorithms in terms of average QoE and fairness.

In recent years, 5G-V2X has promoted the advancement of autonomous vehicles, enabling the latter to obtain more information via 5G networks. However, fast-moving vehicles have to perform frequent handover authentication with base stations in vulnerable wireless channels, which can cause access failures and affect smooth driving. The digital twin is the virtual agent in cyberspace to reliably provide real-time decisions and added-value services to improve the quality of communication for vehicles by analyzing raw data and interacting with the 5G core network. Based on the capabilities of digital twin, in this paper, we propose digital twin-assisted handover authentication scheme that uses the digital twin as the bridge to exchange necessary parameters in 5G-V2X, thereby intelligently assisting in completing mutual authentication and key negotiation between the vehicle and the target base station in advance and reducing the complexity of the handover process. Furthermore, the security and performance analysis demonstrates that our proposed scheme is secure and efficient.

Binary silhouettes and keypoint-based skeletons have dominated human gait recognition studies for decades since they are easy to extract from video frames. Despite their success in gait recognition for in-the-lab environments, they usually fail in real-world scenarios due to their low information entropy for gait representations. To achieve accurate gait recognition in the wild, this paper presents a novel gait representation, named Gait Parsing Sequence (GPS). GPSs are sequences of fine-grained human segmentation, i.e., human parsing, extracted from video frames, so they have much higher information entropy to encode the shapes and dynamics of fine-grained human parts during walking. Moreover, to effectively explore the capability of the GPS representation, we propose a novel human parsing-based gait recognition framework, named ParsingGait. ParsingGait contains a Convolutional Neural Network (CNN)-based backbone and two light-weighted heads. The first head extracts global semantic features from GPSs, while the other one learns mutual information of part-level features through Graph Convolutional Networks to model the detailed dynamics of human walking. Furthermore, due to the lack of suitable datasets, we build the first parsing-based dataset for gait recognition in the wild, named Gait3D-Parsing, by extending the large-scale and challenging Gait3D dataset. Based on Gait3D-Parsing, we comprehensively evaluate our method and existing gait recognition methods. Specifically, ParsingGait achieves a 17.5% Rank-1 increase compared with the state-of-the-art silhouette-based method. In addition, by replacing silhouettes with GPSs, current gait recognition methods achieve about 12.5% ~ 19.2% improvements in Rank-1 accuracy. The experimental results show a significant improvement in accuracy brought by the GPS representation and the superiority of ParsingGait.

Federated learning (FL) is capable of using the local data sets from large-scale nodes for distributed model training. In FL tasks, training and updating are usually repeated ping-pong processes, in that the model training process between devices (workers) and task publisher (TP) needs to be repeated for multiple rounds towards the global model convergence. However, a worker is typically selfish to save its local resource, and even with an incentive mechanism in place at the beginning of model training, a worker may not be honest to participate in all training rounds, leading to poor performance in global model convergence. To enable long-term cooperation in FL, however, has rarely been considered in the existing literature which motivates our work. In this paper, the multi-round FL is modeled as a dynamic repeated game. To exploit the long-term cooperation gain, a general trigger strategy is deployed as the punishment for free-riding and the Nash equilibrium (NE) of the repeated game is derived. Based on the game theoretic analysis, we develop a NE-driven incentive mechanism to guide the TP selects the most effective wages to motivate workers towards long-term cooperation and avoid midway free-riding. Simulation results show the effectiveness of our proposal.

A fundamental issue of the vehicular digital twin (DT) is efficiently synchronizing the data between the DT and the vehicular user (VUE). In this paper, we consider the heterogeneous vehicular networks (HetVNets) in which a VUE can connect to the network through different networks. The HetVNets can improve the efficiency of communication by providing seamless connections. However, the uneven distribution of VUEs and the dynamics of HetVNets make the environment more complex. Therefore, we propose the network selection algorithm for data synchronization between VUEs and DTs in the HetVNets, where the behaviour between the VUEs is considered as a competition for wireless resources. A learning-based prediction model residing in the DT is developed where the DT can predict the waiting time of each relay and transmit the predicted results to the VUE for decision-making. We model the network selection problem as a potential game considering both the transmission time and the waiting time obtained from the prediction model and prove the existence of Nash equilibrium (NE). We analyze the performance of the proposed algorithm, and simulation results show that our approach can effectively find the optimal strategy while achieving a fast convergence speed and high-level performance compared to the baselines.

Existing studies for gait recognition are dominated by 2D representations like the silhouette or skeleton of the human body in constrained scenes. However, humans live and walk in the unconstrained 3D space, so projecting the 3D human body onto the 2D plane will discard a lot of crucial information like the viewpoint, shape, and dynamics for gait recognition. Therefore, this paper aims to explore dense 3D representations for gait recognition in the wild, which is a practical yet neglected problem. In particular, we propose a novel framework to explore the 3D Skinned Multi-Person Linear (SMPL) model of the human body for gait recognition, named SMPLGait. Our framework has two elaborately-designed branches of which one extracts appearance features from silhouettes, the other learns knowledge of 3D viewpoints and shapes from the 3D SMPL model. In addition, due to the lack of suitable datasets, we build the first large-scale 3D representation-based gait recognition dataset, named Gait3D. It contains 4,000 subjects and over 25,000 sequences extracted from 39 cameras in an unconstrained indoor scene. More importantly, it provides 3D SMPL models recovered from video frames which can provide dense 3D information of body shape, viewpoint, and dynamics. Based on Gait3D, we comprehensively compare our method with existing gait recognition approaches, which reflects the superior performance of our framework and the potential of 3D representations for gait recognition in the wild. The code and dataset are available at: https://gait3d.github.io.

This paper considers the downlink data synchronization from the digital twin (DT) to the vehicle, in which a vehicle drives through consecutive roadside units (RSUs) along its trip, and the DT on the cloud transmits the data to the vehicle through the relay of RSUs. To this goal, the DT first chops the data into blocks and cache them in the RSUs along the driving path of the vehicle. The vehicle can then retrieve the blocks when driving into the RSU's coverage to recover the data. Since RSUs have different cache capacities and communication costs, the DT needs to determine how to optimally distribute the data blocks at RSUs so that vehicles can finish downloading all the data before the deadline yet with the minimal cost. To determine the optimal delivery strategy of DT, we model the problem as an optimization framework subject to the time-varying wireless channel of RSUs, their service load and the communication cost. We then resort to the Lyapunov optimization to derive a distributed solution. Using extensive simulation results, we demonstrate that our scheme can effectively reduce the cost of data synchronization and improve the network load performance.

Existing studies for gait recognition are dominated by in-the-lab scenarios. Since people live in real-world senses, gait recognition in the wild is a more practical problem that has recently attracted the attention of the community of multimedia and computer vision. Current methods that obtain state-of-the-art performance on in-the-lab benchmarks achieve much worse accuracy on the recently proposed in-the-wild datasets because these methods can hardly model the varied temporal dynamics of gait sequences in unconstrained scenes. Therefore, this paper presents a novel multi-hop temporal switch method to achieve effective temporal modeling of gait patterns in real-world scenes. Concretely, we design a novel gait recognition network, named Multi-hop Temporal Switch Network (MTSGait), to learn spatial features and multi-scale temporal features simultaneously. Different from existing methods that use 3D convolutions for temporal modeling, our MTSGait models the temporal dynamics of gait sequences by 2D convolutions. By this means, it achieves high efficiency with fewer model parameters and reduces the difficulty in optimization compared with 3D convolution-based models. Based on the specific design of the 2D convolution kernels, our method can eliminate the misalignment of features among adjacent frames. In addition, a new sampling strategy, i.e., non-cyclic continuous sampling, is proposed to make the model learn more robust temporal features. Finally, the proposed method achieves superior performance on two public gait in-the-wild datasets, i.e., GREW and Gait3D, compared with state-of-the-art methods.

This work proposes AMIS, an edge computing-based adaptive video streaming system. AMIS explores the power of edge computing in three aspects. First, with video contents pre-cached in the local buffer, AMIS is content-aware which adapts the video playout strategy based on the scene features of video contents and quality of experience (QoE) of users. Second, AMIS is channel-aware which measures the channel conditions in real-time and estimates the wireless bandwidth. Third, by integrating the content features and channel estimation, AMIS applies the deep reinforcement learning model to optimize the playout strategy towards the best QoE. Therefore, AMIS is an intelligent content- and channel-aware scheme which fully explores the intelligence of edge computing and adapts to general environments and QoE requirements. Using trace-driven simulations, we show that AMIS can succeed in improving the average QoE by 14%-46% as compared to the state-of-the-art adaptive bitrate algorithms.

Gait, i.e., the movement pattern of human limbs during locomotion, is a promising biometrie for identification of persons. Despite significant improvement in gait recognition with deep learning, existing studies still neglect a more practical but challenging scenario - unsupervised cross-domain gait recognition which aims to learn a model on a labeled dataset then adapt it to an unlabeled dataset. Due to the domain shift and class gap, directly applying a model trained on one source dataset to other target datasets usually obtains very poor results. Therefore, this paper proposes a Transferable Neighborhood Discovery (TraND) framework to bridge the domain gap for unsupervised cross-domain gait recognition. To learn effective prior knowledge for gait representation, we first adopt a backbone network pre- trained on the labeled source data in a supervised manner. Then we design an end-to-end trainable approach to automatically discover the confident neighborhoods of unlabeled samples in the latent space. During training, the class consistency indicator is adopted to select confident neighborhoods of samples based on their entropy measurements. Moreover, we explore a high- entropy-first neighbor selection strategy, which can effectively transfer prior knowledge to the target domain. Our method achieves the state-of-the-art results on two public datasets, i.e., CASIA-B and OU-LP.

Vehicle re-identification (Re-Id) is a challenging task due to the inter-class similarity, the intra-class difference, and the cross-view misalignment of vehicle parts. Although recent methods achieve great improvement by learning detailed features from keypoints or bounding boxes of parts, vehicle Re-Id is still far from being solved. Different from existing methods, we propose a Parsing-guided Cross-part Reasoning Network, named as PCRNet, for vehicle Re-Id. The PCRNet explores vehicle parsing to learn discriminative part-level features, model the correlation among vehicle parts, and achieve precise part alignment for vehicle Re-Id. To accurately segment vehicle parts, we first build a large-scale Multi-grained Vehicle Parsing (MVP) dataset from surveillance images. With the parsed parts, we extract regional features for each part and build a part-neighboring graph to explicitly model the correlation among parts. Then, the graph convolutional networks (GCNs) are adopted to propagate local information among parts, which can discover the most effective local features of varied viewpoints. Moreover, we propose a self-supervised part prediction loss to make the GCNs generate features of invisible parts from visible parts under different viewpoints. By this means, the same vehicle from different viewpoints can be matched with the well-aligned and robust feature representations. Through extensive experiments, our PCRNet significantly outperforms the state-of-the-art methods on three large-scale vehicle Re-Id datasets.