Author name cluster

Jinlong Wang

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

5 papers

1 author row

AAAI Conference 2025 Conference Paper

A²RNet: Adversarial Attack Resilient Network for Robust Infrared and Visible Image Fusion

Jiawei Li
Hongwei Yu
Jiansheng Chen
Xinlong Ding
Jinlong Wang
Jinyuan Liu
Bochao Zou
Huimin Ma

Infrared and visible image fusion (IVIF) is a crucial technique for enhancing visual performance by integrating unique information from different modalities into one fused image. Exiting methods pay more attention to conducting fusion with undisturbed data, while overlooking the impact of deliberate interference on the effectiveness of fusion results. To investigate the robustness of fusion models, in this paper, we propose a novel adversarial attack resilient network, called A2RNet. Specifically, we develop an adversarial paradigm with an anti-attack loss function to implement adversarial attacks and training. It is constructed based on the intrinsic nature of IVIF and provide a robust foundation for future research advancements. We adopt a Unet as the pipeline with a transformer-based defensive refinement module (DRM) under this paradigm, which guarantees fused image quality in a robust coarse-to-fine manner. Compared to previous works, our method mitigates the adverse effects of adversarial perturbations, consistently maintaining high-fidelity fusion results. Furthermore, the performance of downstream tasks can also be well maintained under adversarial attacks.

PDF Details DOI

EAAI Journal 2025 Journal Article

Dynamic scheduling in flexible and hybrid disassembly systems with manual and automated workstations using reward-shaping enhanced reinforcement learning

Jinlong Wang
Qihuiyang Liang
Min Li
Zelin Qu
Yuanyuan Zhang

As e-waste grows at an alarming rate, efficient disassembly systems have become crucial for sustainable production practices. Existing disassembly systems rely heavily on fixed automation and limited manual intervention, making it challenging to adapt to dynamic issues such as workstation failures and system bottlenecks, leading to inefficiencies and suboptimal resource allocation. To address these issues, a hybrid disassembly system is developed that integrates manual and automated workstations, allowing for the flexible variation of manual resources as needed to optimize the disassembly process, with a focus on reducing time and maximizing profit. Through the proposal of a Proximal Policy Optimization (PPO) algorithm enhanced with Reward-Shaping, the research effectively tackles key challenges of uncertainty and dynamic conditions in disassembly systems, including workstation failures and system bottlenecks. These issues are explored through a refrigerator disassembly simulation model. The results demonstrate that the PPO algorithm significantly outperforms traditional rule-based methods and two other reinforcement learning techniques in managing complex dynamic scheduling and resource allocation tasks, offering greater efficiency and flexibility. These findings contribute to the advancement of automated disassembly processes and their integration into modern industrial systems.

Details DOI

AAAI Conference 2025 Conference Paper

L4DR: LiDAR-4DRadar Fusion for Weather-Robust 3D Object Detection

Xun Huang
Ziyu Xu
Hai Wu
Jinlong Wang
Qiming Xia
Yan Xia
Jonathan Li
Kyle Gao

LiDAR-based 3D object detection is crucial for autonomous driving. However, due to the quality deterioration of LiDAR point clouds, it suffers from performance degradation in adverse weather conditions. Fusing LiDAR with the weatherrobust 4D radar sensor is expected to solve this problem; however, it faces challenges of significant differences in terms of data quality and the degree of degradation in adverse weather. To address these issues, we introduce L4DR, a weather-robust 3D object detection method that effectively achieves LiDAR and 4D Radar fusion. Our L4DR proposes Multi-Modal Encoding (MME) and Foreground-Aware Denoising (FAD) modules to reconcile sensor gaps, which is the first exploration of the complementarity of early fusion between LiDAR and 4D radar. Additionally, we design an Inter-Modal and IntraModal ({IM}2) parallel feature extraction backbone coupled with a Multi-Scale Gated Fusion (MSGF) module to counteract the varying degrees of sensor degradation under adverse weather conditions. Experimental evaluation on a VoD dataset with simulated fog proves that L4DR is more adaptable to changing weather conditions. It delivers a significant performance increase under different fog levels, improving the 3D mAP by up to 20.0% over the traditional LiDAR-only approach. Moreover, the results on the K-Radar dataset validate the consistent performance improvement of L4DR in realworld adverse weather conditions.

PDF Details DOI

EAAI Journal 2025 Journal Article

Multi-echelon inventory optimization of waste electrical and electronic equipment closed-loop supply chain based on reinforcement learning under carbon tax policy

Jinlong Wang
Shangzhuo Zhou
Min Li
Guanyu Ren
Xianquan Ren
Xiaoyun Xiong
Yuanyuan Zhang

In response to environmental challenges posed by waste electrical and electronic equipment (WEEE), the WEEE closed-loop supply chain (CLSC) has emerged as a crucial means to promote circular economy through the recycling and reuse of WEEE, which not only reduces waste emissions but also improves the efficiency of resource utilization. In practice, both economic and environmental benefits must be considered in sustainable manufacturing within the CLSC to ensure the sustainable development of enterprises. Therefore, a multi-echelon, multi-period inventory model for the CLSC under carbon tax policy is developed in this paper, which innovatively introduces the carbon footprint to assess environmental impact and integrates the impacts of collection planning and the uncertainty of recycling quantity and product demand on the system, aiming to minimize total enterprise costs. To address the uncertainties in the model, the Proximal Policy Optimization (PPO) algorithm is employed to train a reinforcement learning (RL) agent. This agent enables enterprises to dynamically adjust internal strategies such as collection and production, as well as external procurement strategies, based on inventory levels and market conditions. By internalizing carbon emission costs through the carbon tax rate, the RL agent optimizes total costs while achieving a balance between economic and environmental benefits. Numerical experiments demonstrate that the PPO algorithm outperforms the traditional inventory management policy in terms of both cost control and carbon footprint reduction. Moreover, the moderate carbon tax policy on the CLSC appropriately increases the cost of enterprises while significantly reducing their carbon footprint and promoting sustainable development.

Details DOI

EAAI Journal 2025 Journal Article

Risk identification of listed companies violation by integrating knowledge graph and multi-source risk factors

Jinlong Wang
Pengjun Li
Yingmin Liu
Xiaoyun Xiong
Yuanyuan Zhang
Zhihan Lv

The regulatory compliance supervision of listed enterprises is of great significance for ensuring the stable operation of financial markets. However, existing graph propagation algorithms used to identify corporate violations are limited by their inherent randomness. At the same time, current methods consider a relatively narrow range of risk dimensions, making it difficult to accurately distinguish the different characteristics of enterprises. To this end, this paper designs a Propagation of Violation Risks for Listed Enterprises (PVR-LE) algorithm, which reduces the risk of false propagation through a risk weight decay mechanism and dynamic updating of risk propagation patterns. Furthermore, a Multi-source Risk Fusion Neural Network (MRFNN) for corporate violation prediction tasks is proposed, which fuses the liquidity risk characteristics between enterprises, clustering characteristics, and the enterprise’s risk characteristics to endow the violating enterprise nodes with more distinctive and characteristic vector features, thereby identifying whether there are violations in the company. At the same time, a generative adversarial network is used to generate samples of violating enterprises to solve the problem of class imbalance. Experiments are conducted on a real dataset constructed from information on Chinese listed companies, and this method improves the accuracy, recall, the weighted harmonic mean of precision and recall(F1-score), and geometric mean(G-mean) metrics by 2. 12%, 3. 19%, 2. 14%, and 2. 79%, respectively, compared to the best performance of existing models. The experimental results show that the proposed method effectively improves the accuracy of identifying corporate violations and helps promote the development of intelligent regulatory work for listed enterprises.

Details DOI