16 March 2022 in IEEE Access
The visible-infrared pedestrian re- identification (VI Re-ID) task aims to match cross-modality pedestrian images with the same labels. Most current methods focus on mitigating the modality discrepancy by adopting a two-stream network and identity supervision. Based on current methods, we propose a novel feature fusion and center aggregation learning network ( $F^{2}$ CALNet) for cross-modality pedestrian re- identification. $F^{2}$ CALNet focuses on learning modality-irrelevant features by simultaneously reducing inter-modality discrepancies and increasing the inter-identity variations in a single framework. Specifically, we first adopt a two-stream backbone network to extract modality-independent and modality-shared information. Then, we embed modality mitigation modules in a two-stream network to learn feature maps that are stripped of the modality information. Finally, we devise a feature fusion and center aggregation learning module, which first merges two different granularity features to learn distinguishing features, then, we organize two kinds of center-based loss functions to reduce the intra-identity inter- and intra-modality differences and increase inter-identity variations by simultaneously pulling the features of the same identity close to their centers and pushing far away the centers of different identities. Extensive experiments on two public cross-modality datasets (SYSU-MM01 and RegDB) show that $F^{2}$ CALNet is superior to the state-of-the-art approaches. Furthermore, on the SYSU-MM01 datasets, our model outperforms the baseline by 5.52% and 4.25% for the accuracy of rank1 and mAP, respectively.
15 July 2003 in Communications in Theoretical Physics
15 June 2003 in Communications in Theoretical Physics
14 November 2022 in IEEE Access
Visible-Infrared person re-identification (VI Re-ID) is a challenging but practical task that aims at matching pedestrian images between the visible(daytime) modality and the infrared(nighttime) modality, playing an important role in criminal investigation and intelligent video surveillance applications. Numerous previous studies focused on alleviating the modality discrepancy and obtaining discriminating features by devising complex networks for VI Re-ID, but a cumbersome network structure is not suitable for practical industrial applications. In this paper, we propose a novel fusion method of global and part features to extract distinguishing features and alleviate cross-modality differences, named Global and Part Feature Fusion network(GPFF), which has not been well studied in the current literature. Specifically, we first adopt a dual-stream ResNet50 as a backbone network to alleviate the modality discrepancy. Then, We explore how to fuse global and local features to obtain discriminative features. Finally, we apply a heterogeneous center triplet loss(hetero-center triplet loss) instead of traditional triplet loss to guide sample center learning. Our proposed approach is simple but effective, and can remarkably boost the performance of VI Re-ID. The results of experiments on two public datasets(SYSU-MM01 and RegDB) demonstrate that our approach is superior to the state-of-the-art methods. Through experiments, we find that the effective fusion of global features and local features plays an important role in extracting discriminative features.
1 January 2023 in Fractals
The development of 5G technology has enabled the cloud-internet of things (IoT) to impact all areas of our lives. Sensors in cloud-IoT generate large-scale data, and the demand for massive data processing is also increasing. The performance of a single machine can no longer meet the needs of existing users. In contrast, a data center (DC) integrates computing power and storage resources through a specific network topology and satisfies the need to process massive data. Regarding large-scale heterogeneous traffic in DCs, differentiated traffic scheduling on demand reduces transmission latency and improves throughput. Therefore, this paper presents a traffic scheduling method based on deep Q-networks (DQN). This method collects network parameters, delivers them to the environment module, and completes the environment construction of network information and reinforcement learning elements through the environment module. Thus, the final transmission path of the elephant flow is converted based on the action given by DQN. The experimental results show that the method proposed in this paper effectively reduces the transmission latency and improves the link utilization and throughput to a certain extent.
17 December 2025 in IEEE Internet of Things Journal
Emergency medical transportation systems require intelligent network management where high-level user intents that can be automatically translated into operational policies. To address this need within intent-based networking frameworks for intelligent transportation systems, we propose SDAETC-Med, a novel architecture combining stacked denoising auto-encoders (DAEs), transformers, and convolutional long short-term memory networks for medical transport trajectory classification. Our approach enables intent-based automation by providing accurate vehicle type identification that supports policy refinement and network resource allocation. The architecture addresses three core challenges: reducing noise in GPS trajectory data from emergency vehicles, extracting temporal patterns across different emergency response phases, and capturing spatiotemporal features of various medical transport modes. By integrating with intent-based networking architectures, SDAETC-Med enables natural language intent expression, such as “prioritize ambulance connectivity during emergency responses,” to be automatically translated into network policies through accurate vehicle classification. Experiments on GeoLife and Sussex–Huawei Locomotion (SHL) datasets demonstrate that SDAETC-Med achieves superior classification accuracy (92.3% and 92.8%, respectively), outperforming six baseline methods by margins of 0.4–24.1 percentage points. The framework provides a foundation for intent-driven medical resource scheduling and route optimization within intelligent transportation systems, ultimately enabling autonomous network management that improves patient outcomes in time-critical emergency scenarios.