Pub Date : 2025-01-14DOI: 10.1016/j.compeleceng.2025.110063
Mohammad Sarkhosh, Abbas Fattahi
In today's smart grid, the proliferation of distributed generators (DGs) and electric vehicles (EVs) underscores the importance of coordinating their activities. This coordination aims to leverage these resources to enhance network efficiency and mitigate the risks associated with uncoordinated actions, such as charging during peak times, which can have undesirable consequences on grid stability and reliability. To maintain the privacy of agents and lessen their computational workload, we propose the proximal-tracking distributed optimization algorithm (PTDOA) aimed at minimizing the overall operation cost by coordinating agents, including DGs and electric vehicle aggregators (EVAs). PTDOA enables agents to coordinate and optimize their operations independently. Finally, the proposed approach is evaluated using a 33-bus distribution test network containing EVAs and DGs. The results showcase that the proposed algorithm effectively maximizes agent profits while meeting demand requirements and maintaining bus voltage profiles within standard values.
{"title":"A fully distributed model for coordinated operation of distribution generators and electric vehicle aggregators","authors":"Mohammad Sarkhosh, Abbas Fattahi","doi":"10.1016/j.compeleceng.2025.110063","DOIUrl":"10.1016/j.compeleceng.2025.110063","url":null,"abstract":"<div><div>In today's smart grid, the proliferation of distributed generators (DGs) and electric vehicles (EVs) underscores the importance of coordinating their activities. This coordination aims to leverage these resources to enhance network efficiency and mitigate the risks associated with uncoordinated actions, such as charging during peak times, which can have undesirable consequences on grid stability and reliability. To maintain the privacy of agents and lessen their computational workload, we propose the proximal-tracking distributed optimization algorithm (PTDOA) aimed at minimizing the overall operation cost by coordinating agents, including DGs and electric vehicle aggregators (EVAs). PTDOA enables agents to coordinate and optimize their operations independently. Finally, the proposed approach is evaluated using a 33-bus distribution test network containing EVAs and DGs. The results showcase that the proposed algorithm effectively maximizes agent profits while meeting demand requirements and maintaining bus voltage profiles within standard values.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110063"},"PeriodicalIF":4.0,"publicationDate":"2025-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143144610","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-13DOI: 10.1016/j.compeleceng.2024.110052
Vishalakshi, Shobha Rani, Hanumantharaju
Image and video enhancement under low-light conditions is challenging, as the task involves more than just brightness adjustment. Without addressing issues such as artifacts, distortions, and noise in dark regions, brightness improvement alone can worsen the quality. This paper presents a novel approach to low-light image and video enhancement based on the adaptive fusion of Dual Super-Resolution Generative Adversarial Network (DSRGAN) models, followed by Top-Hat Gradient-Domain Filtering (THGDF). A soft thresholding mechanism is used to integrate the Memory Residual Super-Resolution Generative Adversarial Network (MRSRGAN) and the Weighted Perception Super-Resolution Generative Adversarial Network (WPSRGAN). MRSRGAN enhances fine details, improving the objective performance of the image, while WPSRGAN improves overall details, enhancing the subjective performance. Top-hat gradient-domain filtering is then applied to remove artifacts, distortions, and noise in both images and videos, resulting in outstanding perception scores. The proposed approach is validated using the quality assessment metrics such as Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measurement (SSIM), and Information Fidelity Criterion (IFC). Extensive experiments conducted on publicly available source codes and databases demonstrate that the proposed method is more effective than the existing state-of-the-art techniques.
{"title":"A novel approach to low-light image and video enhancement using adaptive dual super-resolution generative adversarial networks and top-hat filtering","authors":"Vishalakshi, Shobha Rani, Hanumantharaju","doi":"10.1016/j.compeleceng.2024.110052","DOIUrl":"10.1016/j.compeleceng.2024.110052","url":null,"abstract":"<div><div>Image and video enhancement under low-light conditions is challenging, as the task involves more than just brightness adjustment. Without addressing issues such as artifacts, distortions, and noise in dark regions, brightness improvement alone can worsen the quality. This paper presents a novel approach to low-light image and video enhancement based on the adaptive fusion of Dual Super-Resolution Generative Adversarial Network (DSRGAN) models, followed by Top-Hat Gradient-Domain Filtering (THGDF). A soft thresholding mechanism is used to integrate the Memory Residual Super-Resolution Generative Adversarial Network (MRSRGAN) and the Weighted Perception Super-Resolution Generative Adversarial Network (WPSRGAN). MRSRGAN enhances fine details, improving the objective performance of the image, while WPSRGAN improves overall details, enhancing the subjective performance. Top-hat gradient-domain filtering is then applied to remove artifacts, distortions, and noise in both images and videos, resulting in outstanding perception scores. The proposed approach is validated using the quality assessment metrics such as Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measurement (SSIM), and Information Fidelity Criterion (IFC). Extensive experiments conducted on publicly available source codes and databases demonstrate that the proposed method is more effective than the existing state-of-the-art techniques.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110052"},"PeriodicalIF":4.0,"publicationDate":"2025-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143144609","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-12DOI: 10.1016/j.compeleceng.2024.110054
Duygu Nazan Gençoğlan
This study presents a novel hybrid reconfigurable antenna design optimized using an Adaptive Neuro-Fuzzy Inference System (ANFIS) enhanced with a Simulated Annealing (SA) algorithm for l-band, C-band, 5G, and ISM applications. The antenna is fabricated on an FR-4 substrate with dimensions of 17 × 28 × 1.6 mm³, and two PIN diodes are employed to achieve frequency and radiation pattern reconfigurability. In the ONON state, the antenna operates in dual bands, covering 1.33–1.38 GHz (L-band) and 3.57–3.95 GHz (C-band). For the OFF-ON state, it operates from 3.56 to 3.95 GHz (C-band, 5G). In the ONOFF state, it covers 1.50–1.54 GHz (L-band) and 5.66–5.90 GHz (ISM band), while in the OFF-OFF state, it operates from 5.49 to 5.82 GHz (ISM band). The antenna exhibits common bands at 3.8 GHz (C-band) and 5.8 GHz (ISM) across different states, facilitating pattern reconfigurability. ANFIS-SA is applied to optimize the switch locations, significantly improving resonance frequency and S11 performance. The antenna supports beam steering between 0° and 180°, enhancing adaptive coverage for modern applications such as Wi-Fi, Vehicle-to-Vehicle (V2 V), and Vehicle-to-Infrastructure (V2I) communication. This study addresses a critical gap by combining hybrid optimization techniques to improve frequency agility and radiation pattern control for next-generation wireless systems.
{"title":"ANFIS-SA-based design of a hybrid reconfigurable antenna for l-Band, C-band, 5G and ISM band applications","authors":"Duygu Nazan Gençoğlan","doi":"10.1016/j.compeleceng.2024.110054","DOIUrl":"10.1016/j.compeleceng.2024.110054","url":null,"abstract":"<div><div>This study presents a novel hybrid reconfigurable antenna design optimized using an Adaptive Neuro-Fuzzy Inference System (ANFIS) enhanced with a Simulated Annealing (SA) algorithm for <span>l</span>-band, C-band, 5G, and ISM applications. The antenna is fabricated on an FR-4 substrate with dimensions of 17 × 28 × 1.6 mm³, and two PIN diodes are employed to achieve frequency and radiation pattern reconfigurability. In the ON<img>ON state, the antenna operates in dual bands, covering 1.33–1.38 GHz (L-band) and 3.57–3.95 GHz (C-band). For the OFF-ON state, it operates from 3.56 to 3.95 GHz (C-band, 5G). In the ON<img>OFF state, it covers 1.50–1.54 GHz (L-band) and 5.66–5.90 GHz (ISM band), while in the OFF-OFF state, it operates from 5.49 to 5.82 GHz (ISM band). The antenna exhibits common bands at 3.8 GHz (C-band) and 5.8 GHz (ISM) across different states, facilitating pattern reconfigurability. ANFIS-SA is applied to optimize the switch locations, significantly improving resonance frequency and S11 performance. The antenna supports beam steering between 0° and 180°, enhancing adaptive coverage for modern applications such as Wi-Fi, Vehicle-to-Vehicle (V2 V), and Vehicle-to-Infrastructure (V2I) communication. This study addresses a critical gap by combining hybrid optimization techniques to improve frequency agility and radiation pattern control for next-generation wireless systems.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110054"},"PeriodicalIF":4.0,"publicationDate":"2025-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143144598","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
One vital challenge of diesel generators is their ability to provide a stable power supply in a weak grid under fault scenarios. This paper, therefore, proposes a novel reinforcement learning strategy to enhance the reliability of diesel generators via excitation control. The proposed smoothed deep Q-network (SDQN) controller provides continuous action signals and, therefore, potentially resolves the drawbacks of the conventional deep Q-network (DQN) in maximising cumulative reward due to restricted action space. The hyper-parameters of the SDQN-based learning controller are selected using the Taguchi method, which guarantees convergence of the controller and thereby ensures maximisation of the reward function. In order to provide continuous action signals, a particle swarm optimisation (PSO)-based smoothing procedure is carried out. The advantage of the proposed controller is studied under various conditions, including tripping and fault scenarios. The effectiveness of the controller is compared with other RL agents with continuous action space and also traditional power system stabilisers (PSS). The simulation results demonstrate that the performance of SDQN is superior to that of other controllers in regulating the terminal voltage and rotor angle under various conditions of diesel engine generator operations.
{"title":"Design of smoothed deep Q-network for excitation control of grid-tied diesel generator","authors":"Gunawan Dewantoro , Faizal Hafiz , Akshya Swain , Nitish Patel","doi":"10.1016/j.compeleceng.2025.110058","DOIUrl":"10.1016/j.compeleceng.2025.110058","url":null,"abstract":"<div><div>One vital challenge of diesel generators is their ability to provide a stable power supply in a weak grid under fault scenarios. This paper, therefore, proposes a novel reinforcement learning strategy to enhance the reliability of diesel generators via excitation control. The proposed smoothed deep Q-network (SDQN) controller provides continuous action signals and, therefore, potentially resolves the drawbacks of the conventional deep Q-network (DQN) in maximising cumulative reward due to restricted action space. The hyper-parameters of the SDQN-based learning controller are selected using the Taguchi method, which guarantees convergence of the controller and thereby ensures maximisation of the reward function. In order to provide continuous action signals, a particle swarm optimisation (PSO)-based smoothing procedure is carried out. The advantage of the proposed controller is studied under various conditions, including tripping and fault scenarios. The effectiveness of the controller is compared with other RL agents with continuous action space and also traditional power system stabilisers (PSS). The simulation results demonstrate that the performance of SDQN is superior to that of other controllers in regulating the terminal voltage and rotor angle under various conditions of diesel engine generator operations.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110058"},"PeriodicalIF":4.0,"publicationDate":"2025-01-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143144611","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-10DOI: 10.1016/j.compeleceng.2024.110029
Chuan Lu, Senlin Luo, Limin Pan
Spectre attacks can steal private information via some usual code snippets, and the difficulty in detecting Spectre gadgets is to mine the differences between the same code snippet in different context states. However, existing methods manually construct global behavior features of Spectre gadgets, which cannot capture fine-grained data flow and control flow between codes to distinguish normal code snippets with similar features, resulting in low detection accuracy. Meanwhile, existing methods select the speculative execution path for testing based on the judgment result of conditional instruction, which is prone to cause a path being repeatedly detected under similar conditions, leading to inefficient detection. Therefore, a Spectre gadget detection method with Strengthened Context Relevance (SCR-Spectre) is proposed. SCR-Spectre presents a strengthened context relevance SCR-BERT, which learns fine-grained data flow and control flow by the context relevance, increasing the detection precision. Furthermore, a novel combined dynamic and static testing framework is pioneered, which dynamically screens conditional instruction and statically extracts features of all speculative execution paths of this instruction, avoiding repetitive testing. Experimental results show that SCR-Spectre significantly outperforms state-of-the-art correlation methods. This method fuses fine-grained features between code contexts, strengthening the distinguishability of Spectre gadgets.
{"title":"SCR-Spectre: Spectre gadget detection method with strengthened context relevance","authors":"Chuan Lu, Senlin Luo, Limin Pan","doi":"10.1016/j.compeleceng.2024.110029","DOIUrl":"10.1016/j.compeleceng.2024.110029","url":null,"abstract":"<div><div>Spectre attacks can steal private information via some usual code snippets, and the difficulty in detecting Spectre gadgets is to mine the differences between the same code snippet in different context states. However, existing methods manually construct global behavior features of Spectre gadgets, which cannot capture fine-grained data flow and control flow between codes to distinguish normal code snippets with similar features, resulting in low detection accuracy. Meanwhile, existing methods select the speculative execution path for testing based on the judgment result of conditional instruction, which is prone to cause a path being repeatedly detected under similar conditions, leading to inefficient detection. Therefore, a Spectre gadget detection method with <strong><u>S</u></strong>trengthened <strong><u>C</u></strong>ontext <strong><u>R</u></strong>elevance (SCR-Spectre) is proposed. SCR-Spectre presents a strengthened context relevance SCR-BERT, which learns fine-grained data flow and control flow by the context relevance, increasing the detection precision. Furthermore, a novel combined dynamic and static testing framework is pioneered, which dynamically screens conditional instruction and statically extracts features of all speculative execution paths of this instruction, avoiding repetitive testing. Experimental results show that SCR-Spectre significantly outperforms state-of-the-art correlation methods. This method fuses fine-grained features between code contexts, strengthening the distinguishability of Spectre gadgets.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110029"},"PeriodicalIF":4.0,"publicationDate":"2025-01-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143144542","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-10DOI: 10.1016/j.compeleceng.2024.110038
Maocheng Bai, Xiaosheng Yu
Deep learning-based models have obtained great improvements in facial expression recognition (FER). However, these deep models have high computational complexity and more memory during training and inference, limiting their scalability in deploying on portable devices. In addition, the exploration of the intrinsic connection between facial muscle movements and expressions has always been a huge challenge. To resolve these dilemmas, we propose an effective binary tiny machine learning (TinyML) model by combining two different attention mechanisms and binary operations. Specifically, to exploit the muscle movements in different facial expressions, we propose an effective lightweight deep model by introducing channel and spatial attention mechanisms in which learning weights for different regions can enable the network to focus on regions associated with facial expressions. Moreover, we introduce the scale factor-based binary operation to improve the inference speed. Extensive experiments on three public facial expression datasets prove that our proposed model can achieve advanced performance with 70 K parameters and 0.96MB model size. We have ported and tested our model on the Seeed XIAO ESP32S3 Sense platform, showing the superiority of what was proposed in terms of inference speed.
{"title":"Lightweight emotion analysis solution using tiny machine learning for portable devices","authors":"Maocheng Bai, Xiaosheng Yu","doi":"10.1016/j.compeleceng.2024.110038","DOIUrl":"10.1016/j.compeleceng.2024.110038","url":null,"abstract":"<div><div>Deep learning-based models have obtained great improvements in facial expression recognition (FER). However, these deep models have high computational complexity and more memory during training and inference, limiting their scalability in deploying on portable devices. In addition, the exploration of the intrinsic connection between facial muscle movements and expressions has always been a huge challenge. To resolve these dilemmas, we propose an effective binary tiny machine learning (TinyML) model by combining two different attention mechanisms and binary operations. Specifically, to exploit the muscle movements in different facial expressions, we propose an effective lightweight deep model by introducing channel and spatial attention mechanisms in which learning weights for different regions can enable the network to focus on regions associated with facial expressions. Moreover, we introduce the scale factor-based binary operation to improve the inference speed. Extensive experiments on three public facial expression datasets prove that our proposed model can achieve advanced performance with 70 K parameters and 0.96MB model size. We have ported and tested our model on the Seeed XIAO ESP32S3 Sense platform, showing the superiority of what was proposed in terms of inference speed.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110038"},"PeriodicalIF":4.0,"publicationDate":"2025-01-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143145503","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-10DOI: 10.1016/j.compeleceng.2024.110057
Hyunju Kim , Woosang Im , Sooyong Jeong , Hyunil Kim , Changho Seo , Chanku Kang
Isogeny-based cryptography is secure in a quantum computing environment and offers relatively small key sizes compared to other post-quantum cryptographic schemes. However, the CSIDH isogeny-based cryptography scheme is vulnerable to fault injection attacks, particularly the Disorientation Attack. This paper analyzes the Disorientation Attack on CSIDH and proposes an optimized method to recover the secret key for CTIDH, an extension of CSIDH. In CTIDH, the secret key is divided into multiple batches. The sign of each batch determines whether its isogeny operations proceed in a positive or negative direction. We explore the feasibility of adapting this attack to CTIDH to recover the secret keys. However, because of the unique isogeny operation in CTIDH, direct key recovery presents significant challenges. To address this, we propose an optimized Disorientation Attack for CTIDH, utilizing the differences in the number of points across batches. We also present the details and results of the implementation. Additionally, using the recovered secret keys from this optimized attack, we can enhance the existing Disorientation Attack to recover additional key values.
{"title":"Recovery for secret key in CTIDH-512 through Fault Injection Attack","authors":"Hyunju Kim , Woosang Im , Sooyong Jeong , Hyunil Kim , Changho Seo , Chanku Kang","doi":"10.1016/j.compeleceng.2024.110057","DOIUrl":"10.1016/j.compeleceng.2024.110057","url":null,"abstract":"<div><div>Isogeny-based cryptography is secure in a quantum computing environment and offers relatively small key sizes compared to other post-quantum cryptographic schemes. However, the CSIDH isogeny-based cryptography scheme is vulnerable to fault injection attacks, particularly the Disorientation Attack. This paper analyzes the Disorientation Attack on CSIDH and proposes an optimized method to recover the secret key for CTIDH, an extension of CSIDH. In CTIDH, the secret key is divided into multiple batches. The sign of each batch determines whether its isogeny operations proceed in a positive or negative direction. We explore the feasibility of adapting this attack to CTIDH to recover the secret keys. However, because of the unique isogeny operation in CTIDH, direct key recovery presents significant challenges. To address this, we propose an optimized Disorientation Attack for CTIDH, utilizing the differences in the number of points across batches. We also present the details and results of the implementation. Additionally, using the recovered secret keys from this optimized attack, we can enhance the existing Disorientation Attack to recover additional key values.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110057"},"PeriodicalIF":4.0,"publicationDate":"2025-01-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143144612","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-10DOI: 10.1016/j.compeleceng.2024.110043
Ziqi Wang, Yang Li, Rui Zhang, Jiabao Wang, Haoran Cui
Few-shot SAR image classification predicts new labels on SAR (synthetic aperture radar) images using only a few labeled samples. Due to the insufficient training data in few-shot learning, modern methods tend to extract as many features as possible, which ignore the redundancy caused by increasing features. In this paper, we observe that it is important to maintain a balance between the abundance and scarcity of features through feature refinement. Building on this insight, we proposed a Feature Refinement network for few-shot SAR image classification (FRSAR). FRSAR is a novel approach for feature balance through feature refinement, which consists of enrich feature extraction (EFE) block, redundant feature refinement (RFR) block, and key feature localization (KFL) block. The EFE block extracts complementary features to achieve a richer feature representation. The RFR block filters out redundant features through a two-stage reconstruction process. The KFL block further refines the classification features by adaptive computing thresholds. Extensive experiments on MSTAR dataset demonstrate that our FRSAR method can achieve better performance than other methods. For example, our method surpasses existing the state-of-the-art method by a large margin (+7.20/% on the 5-way 1-shot task and +7.22% on the 5-way 5-shot task). We believe the feature refinement framework can serve as a strong baseline for further research in wider communities.
{"title":"More diversity, less redundancy: Feature refinement network for few-shot SAR image classification","authors":"Ziqi Wang, Yang Li, Rui Zhang, Jiabao Wang, Haoran Cui","doi":"10.1016/j.compeleceng.2024.110043","DOIUrl":"10.1016/j.compeleceng.2024.110043","url":null,"abstract":"<div><div>Few-shot SAR image classification predicts new labels on SAR (synthetic aperture radar) images using only a few labeled samples. Due to the insufficient training data in few-shot learning, modern methods tend to extract as many features as possible, which ignore the redundancy caused by increasing features. In this paper, we observe that it is important to maintain a balance between the abundance and scarcity of features through feature refinement. Building on this insight, we proposed a <strong>F</strong>eature <strong>R</strong>efinement network for few-shot <strong>SAR</strong> image classification (FRSAR). FRSAR is a novel approach for feature balance through feature refinement, which consists of enrich feature extraction (EFE) block, redundant feature refinement (RFR) block, and key feature localization (KFL) block. The EFE block extracts complementary features to achieve a richer feature representation. The RFR block filters out redundant features through a two-stage reconstruction process. The KFL block further refines the classification features by adaptive computing thresholds. Extensive experiments on MSTAR dataset demonstrate that our FRSAR method can achieve better performance than other methods. For example, our method surpasses existing the state-of-the-art method by a large margin (+7.20/% on the 5-way 1-shot task and +7.22% on the 5-way 5-shot task). We believe the feature refinement framework can serve as a strong baseline for further research in wider communities.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110043"},"PeriodicalIF":4.0,"publicationDate":"2025-01-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143144607","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-10DOI: 10.1016/j.compeleceng.2024.110023
Erika Pequeno dos Santos , Beatriz Silveira Buss , Mauro Augusto da Rosa , Diego Issicaba
This paper introduces an approach to composite reliability evaluation using sequential Monte Carlo simulation with maximum and minimum loadability analysis. The approach utilizes an estimation of the maximum and minimum loadability of a given system state to avoid linearized power flow (PF) and optimal power flow (OPF) assessments in further states characterized by load transitions. The evaluation of the maximum loadability of a system state is also employed within the cross-entropy (CE) optimization procedure, as an alternative performance function to sampled states. Numerical results highlight the effectiveness of the approach and gains in runtime, in comparison with implementations of benchmark methods, for the IEEE RTS 79 test system.
{"title":"Composite reliability evaluation using sequential Monte Carlo simulation with maximum and minimum loadability analysis","authors":"Erika Pequeno dos Santos , Beatriz Silveira Buss , Mauro Augusto da Rosa , Diego Issicaba","doi":"10.1016/j.compeleceng.2024.110023","DOIUrl":"10.1016/j.compeleceng.2024.110023","url":null,"abstract":"<div><div>This paper introduces an approach to composite reliability evaluation using sequential Monte Carlo simulation with maximum and minimum loadability analysis. The approach utilizes an estimation of the maximum and minimum loadability of a given system state to avoid linearized power flow (PF) and optimal power flow (OPF) assessments in further states characterized by load transitions. The evaluation of the maximum loadability of a system state is also employed within the cross-entropy (CE) optimization procedure, as an alternative performance function to sampled states. Numerical results highlight the effectiveness of the approach and gains in runtime, in comparison with implementations of benchmark methods, for the IEEE RTS 79 test system.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110023"},"PeriodicalIF":4.0,"publicationDate":"2025-01-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143144556","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-09DOI: 10.1016/j.compeleceng.2024.110045
Jiyuan Qiu, Chen Jiang
Freespace detection is one of the key technologies for scene understanding and motion planning in autonomous vehicles. However, current research on freespace detection primarily focuses on obstacles provided by objects above the freespace, such as vehicles, pedestrians, and buildings, while less attention is given to impairments within the freespace, such as potholes, defects, and collapses. Moreover, there is a lack of research on the interpretability of artificial intelligence in freespace detection. In this study, we first construct a large-scale off-road freespace detection dataset with impairments (ORIFD). The dataset comprises a total of 24,000 images representing different weather conditions (day, night, snow, etc.) and terrains (concrete roads, dirt roads, rocky paths, etc.). The impairments include potholes, defects, water puddles, and collapses. In addition to adding semantic labels for freespace and impairments, we also create semantic edge labels to enhance the extraction of scene information. Subsequently, we train a novel semantic guidance network (BSGNet) on this dataset, designed to simultaneously perform freespace detection and semantic edge detection tasks. Our framework consists of a deep extended dual-branch encoder, where one branch aggregates multi-scale semantic features, and the other extracts semantic edge information. We also propose an interactive fusion block (IFB) and a global feature aggregation module (GFAM) to enhance the model's feature representation capabilities. Extensive experiments demonstrate that our model outperforms existing state-of-the-art models, achieving superior performance. Additionally, we employ explainable artificial intelligence (XAI) methods to enhance the trustworthiness of our model and implement a method that combines bird's-eye view with the hybrid A* algorithm for generating effective collision-free paths, further extending the application of our research in autonomous vehicles.
{"title":"A bilateral semantic guidance network for detection of off-road freespace with impairments based on joint semantic segmentation and edge detection","authors":"Jiyuan Qiu, Chen Jiang","doi":"10.1016/j.compeleceng.2024.110045","DOIUrl":"10.1016/j.compeleceng.2024.110045","url":null,"abstract":"<div><div>Freespace detection is one of the key technologies for scene understanding and motion planning in autonomous vehicles. However, current research on freespace detection primarily focuses on obstacles provided by objects above the freespace, such as vehicles, pedestrians, and buildings, while less attention is given to impairments within the freespace, such as potholes, defects, and collapses. Moreover, there is a lack of research on the interpretability of artificial intelligence in freespace detection. In this study, we first construct a large-scale off-road freespace detection dataset with impairments (ORIFD). The dataset comprises a total of 24,000 images representing different weather conditions (day, night, snow, etc.) and terrains (concrete roads, dirt roads, rocky paths, etc.). The impairments include potholes, defects, water puddles, and collapses. In addition to adding semantic labels for freespace and impairments, we also create semantic edge labels to enhance the extraction of scene information. Subsequently, we train a novel semantic guidance network (BSGNet) on this dataset, designed to simultaneously perform freespace detection and semantic edge detection tasks. Our framework consists of a deep extended dual-branch encoder, where one branch aggregates multi-scale semantic features, and the other extracts semantic edge information. We also propose an interactive fusion block (IFB) and a global feature aggregation module (GFAM) to enhance the model's feature representation capabilities. Extensive experiments demonstrate that our model outperforms existing state-of-the-art models, achieving superior performance. Additionally, we employ explainable artificial intelligence (XAI) methods to enhance the trustworthiness of our model and implement a method that combines bird's-eye view with the hybrid A* algorithm for generating effective collision-free paths, further extending the application of our research in autonomous vehicles.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"123 ","pages":"Article 110045"},"PeriodicalIF":4.0,"publicationDate":"2025-01-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143145502","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号