Cognitive Robotics最新文献

{"title":"Improving log anomaly detection via spatial pooling: Combining SPClassifier with ensemble method","authors":"Hironori Uchida ,&nbsp;Keitaro Tominaga ,&nbsp;Hideki Itai ,&nbsp;Yujie Li ,&nbsp;Yoshihisa Nakatoh","doi":"10.1016/j.cogr.2024.10.001","DOIUrl":"10.1016/j.cogr.2024.10.001","url":null,"abstract":"<div><div>In the ever-updating field of software development, new bugs emerge daily, requiring significant time for analysis. As a result, research is being conducted on automating bug resolution using techniques such as anomaly detection through deep learning applied to text logs. This study focuses on anomaly detection using text logs and aims to address current challenges. Specifically, we aim to improve the accuracy of SPClassifier, a robust and lightweight AI model capable of handling dynamic log datasets through ad-hoc learning. We employ three ensemble learning methods to enhance the accuracy of SPClassifier. The method that achieved the greatest improvement was Improved Bagging, which combines the non-overlapping sampling of Pasting with the overlapping sampling of Bagging, resulting in a maximum F1-score improvement of 155 %. Additionally, on certain datasets, the F1-score surpassed that of well-known DNN methods by 130 %. Furthermore, the proposed method demonstrated lower variance compared to DNN methods, indicating its advantage, particularly in environments where datasets frequently fluctuate, such as development fields. These results highlight the clear superiority of the proposed method, which is lightweight in terms of computational resources and supports ad-hoc learning.</div></div>","PeriodicalId":100288,"journal":{"name":"Cognitive Robotics","volume":"4 ","pages":"Pages 217-227"},"PeriodicalIF":0.0,"publicationDate":"2024-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142704755","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Big Data Course Multidimensional Evaluation Model based on Knowledge Graph enhanced Transformer","authors":"Ning Liu,&nbsp;Yeyangyi Xiang,&nbsp;Fei Wang,&nbsp;Shuyu Cao","doi":"10.1016/j.cogr.2024.11.003","DOIUrl":"10.1016/j.cogr.2024.11.003","url":null,"abstract":"<div><div>Based on the positioning of training application-oriented and innovative talents in the field of big data, this article aims to address the current situation where the theoretical system of big data course is not complete, the experimental system is unreasonable, and the assessment indicators are not perfect. A Transformer based “1 + 1 + <em>N</em>” big data course unified system and multidimensional evaluation model is constructed, reforms and practices are carried out in terms of improving the course theoretical system, increasing unit experiments and comprehensive experiment cases, and improving process assessment. The Transformer based multi-dimensional evaluation model of the big data course is proposed to solve the current problems of heavy theory and light practice, heavy standardization assessment and light innovation ability training in the course. The proposed course unified system and multidimensional evaluation model had achieved remarkable results, effectively increasing students’ construction of the big data professional knowledge system, enhancing students’ subjective initiative in learning the course, and significantly improving students’ innovative ability and ability to comprehensively solve practical problems.</div></div>","PeriodicalId":100288,"journal":{"name":"Cognitive Robotics","volume":"4 ","pages":"Pages 237-244"},"PeriodicalIF":0.0,"publicationDate":"2024-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143129865","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"RDSM: Underwater multi-AUV relay deployment and selection mechanism in 3D space","authors":"Yafei Liu ,&nbsp;Na Liu ,&nbsp;Hao Li ,&nbsp;Yi Jiang ,&nbsp;Junwu zhu","doi":"10.1016/j.cogr.2024.11.001","DOIUrl":"10.1016/j.cogr.2024.11.001","url":null,"abstract":"<div><div>Underwater Wireless Sensor Networks (UWSNs) are widely used in naval military field and marine resource exploration. However, challenges such as resource inefficiency and unbalanced energy consumption severely hinder their practical applications. In this paper, we establish a model of underwater multi-hop wireless sensor network with multiple AUVs as relay nodes, which describes the data transmission process within the network. Based on this, an underwater multi-AUV Relay Deployment and Selection Mechanism in 3D space (RDSM) is proposed to achieve efficient underwater networking. Specifically, the RDSM includes the following key components. Firstly, an optimized relay node deployment strategy (RNDS) is used to deploy AUV nodes to effectively ensure network connectivity. Compared with traditional methods, this strategy has unique advantages in considering underwater space characteristics and can better adapt to the complex underwater environment. Secondly, a new utility function is constructed by integrating factors such as throughput, energy consumption, and load. The relay selection strategy based on utility maximization (RSS-UM) is used to select the next-hop relay node. This strategy is innovative in improving relay selection efficiency and optimizing network performance. Finally, in response to the problem of rapid energy consumption of relay nodes close to the base station, a power adjustment scheme is introduced to achieve a balance in node energy consumption, which is of great significance for prolonging network lifetime and improving overall stability. Experimental results show that compared with existing methods, the proposed mechanism achieves high utility and throughput, while maintaining balanced node energy consumption.</div></div>","PeriodicalId":100288,"journal":{"name":"Cognitive Robotics","volume":"4 ","pages":"Pages 204-216"},"PeriodicalIF":0.0,"publicationDate":"2024-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142704754","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"YOLOT: Multi-scale and diverse tire sidewall text region detection based on You-Only-Look-Once(YOLOv5)","authors":"Dehua Liu ,&nbsp;Yongqin Tian ,&nbsp;Yibo Xu ,&nbsp;Wenyi Zhao ,&nbsp;Xipeng Pan ,&nbsp;Xu Ji ,&nbsp;Mu Yang ,&nbsp;Huihua Yang","doi":"10.1016/j.cogr.2024.03.001","DOIUrl":"10.1016/j.cogr.2024.03.001","url":null,"abstract":"<div><p>Driving safety is significant to building a people-oriented and harmonious society, Tires are one of the key components of a vehicle and the character information on the tire sidewall is critical to their storage and usage. However, due to the diverse and differentiated features of typographic fonts, simultaneously extracting comprehensive characteristics is an extremely challenging task. To effectively break through these performance degradation issues, a multi-scale tire sidewall text region detection algorithm based on YOLOv5 is introduced, called YOLOT, which fuses comprehensive feature information in both width and depth directions. In this study, we firstly propose the Width and Depth Awareness (WDA) module in the text region detection field and successfully integrated it with the FPN structure to form the WDA-FPN. The purpose of WDA-FPN is to empower the network to capture multi-scale and multi-shape features in images, thereby augmenting the algorithm’s abstraction and representation of image features and concurrently boosting its robustness and generalization performance. Experimental findings indicate that, compared to the primary algorithm, YOLOT achieves significant improvement in accuracy, providing a higher detection reliability. The dataset and code for the paper are available at: https://github.com/Cloude-dehua/YOLOT.</p></div>","PeriodicalId":100288,"journal":{"name":"Cognitive Robotics","volume":"4 ","pages":"Pages 74-87"},"PeriodicalIF":0.0,"publicationDate":"2024-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S266724132400003X/pdfft?md5=19ce0153cf7a9ea3214d8e7517f90940&pid=1-s2.0-S266724132400003X-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140277459","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Scalable and cohesive swarm control based on reinforcement learning","authors":"Marc-Andrė Blais,&nbsp;Moulay A. Akhloufi","doi":"10.1016/j.cogr.2024.05.003","DOIUrl":"https://doi.org/10.1016/j.cogr.2024.05.003","url":null,"abstract":"<div><p>Unmanned vehicles have seen a significant increase in a wide variety of fields such as for logistics, agriculture and other commercial applications. Controlling swarms of unmanned vehicles is a challenging task that requires complex autonomous control systems. Reinforcement learning has been proposed as a solution to this challenge. We propose an approach based on agent masking to enable a simple Deep Q-Network algorithm to scale on large swarms while training on relatively smaller swarms. We train our approach using multiple swarm sizes and learning rates and compare our results using metrics such as the number of collisions. We also compare the ability of our approach to scale on swarms ranging from five to 25 agents using metrics and visual analysis. Our proposed solution was able to guide a swarm of up to 100 agents to a target while keeping a good swarm cohesion and avoiding collision.</p></div>","PeriodicalId":100288,"journal":{"name":"Cognitive Robotics","volume":"4 ","pages":"Pages 88-103"},"PeriodicalIF":0.0,"publicationDate":"2024-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2667241324000053/pdfft?md5=3db6df465ddd69e88b21f962232e9c5e&pid=1-s2.0-S2667241324000053-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141292146","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Intelligent path planning for cognitive mobile robot based on Dhouib-Matrix-SPP method","authors":"Souhail Dhouib","doi":"10.1016/j.cogr.2024.02.001","DOIUrl":"10.1016/j.cogr.2024.02.001","url":null,"abstract":"<div><p>The Mobile Robot Path Problem looks to find the optimal shortest path from the starting point to the target point with collision-free for a mobile robot. This is a popular issue in robotics and in this paper the environment is considered as static and represented as a bidirectional grid map. Besides, the novel optimal method Dhouib-Matrix-SPP (DM-SPP) is applied to create the optimal shortest path for a mobile robot in a static environment. DM-SPP is a greedy method based on a column row navigation in the distance matrix and characterized by its rapidity to solve sparse graphs. The comparative analysis is conducted by applying DM-SPP on thirteen test cases and comparing its results to the results given by four metaheuristics the Max-Min Ant System, the Ant System with punitive measures, the A* and the Improved Hybrid A*. The outcomes acquired from different scenarios indicate that the proposed DM-SPP method can rapidly outperform the four predefined artificial intelligence methods.</p></div>","PeriodicalId":100288,"journal":{"name":"Cognitive Robotics","volume":"4 ","pages":"Pages 62-73"},"PeriodicalIF":0.0,"publicationDate":"2024-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2667241324000028/pdfft?md5=753a3935e8733e20519f0d68f97e618f&pid=1-s2.0-S2667241324000028-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139966008","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Power inspection UAV task assignment matrix reversal genetic algorithm","authors":"Kai Liu ,&nbsp;Meizhao Liu ,&nbsp;Ming Tang ,&nbsp;Chen Zhang","doi":"10.1016/j.cogr.2024.11.006","DOIUrl":"10.1016/j.cogr.2024.11.006","url":null,"abstract":"<div><div>Traditional manual power inspections are characterized by low efficiency, lengthy processes, and high costs. Existing research on UAV-based power inspections has often overlooked critical factors such as the risk levels of target tasks, the duration of tasks executed by UAVs, and the utility per unit task. To address these gaps, this paper proposes a task allocation method for UAV power inspections based on the Time Window Matrix Reversal Genetic Algorithm (TMGA). Firstly, the proposed cost model accounts for the risk levels of inspection tasks and the impact of low-altitude flight on energy consumption. Secondly, an inspection task allocation model is constructed with the goal of maximizing UAV inspection unit utility. The model is then optimized using two-point crossover and single-point reversal mutation operations, which enhance the UAV unit utility and generate an optimal allocation matrix. The performance of TMGA is evaluated through simulation experiments in three different scenarios, comparing it with existing algorithms. The results show that TMGA outperforms these algorithms in terms of average task time, task completion rate, and unit utility. Specifically, TMGA reduces the average task time by 37% compared to the Cluster Grouping Consensus-base Bundle Algorithm and improves task unit utility by 56.91% compared to the Genetic Algorithm.</div></div>","PeriodicalId":100288,"journal":{"name":"Cognitive Robotics","volume":"4 ","pages":"Pages 245-258"},"PeriodicalIF":0.0,"publicationDate":"2024-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143129866","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Robot assisted knee joint RoM exercise: A PID parallel compensator architecture through impedance estimation","authors":"M. Akhtaruzzaman ,&nbsp;Amir A. Shafie ,&nbsp;Md Raisuddin Khan ,&nbsp;Md Mozasser Rahman","doi":"10.1016/j.cogr.2023.11.003","DOIUrl":"10.1016/j.cogr.2023.11.003","url":null,"abstract":"<div><p>Knee joint rehabilitation exercise refers to a therapeutic procedure of a patient having dysfunctions in certain abilities to move knee joint due to some medical conditions like trauma or paralysis. The exercise is basically a series of repeated assistive physical movements within the range of motion (RoM) of the joint. Reflex action of limbs during RoM exercise causes inappropriate balance of load which may cause secondary injuries, such as damages of muscle or tendon tissues. Establishing correlation between impedance data and limb motions is important to solve this problem. This paper aims to design and modeling of a robotic arm with an original approach in control strategy which is developed based on the correlation in between the joint-impedances and joint-motion characteristics during exercise. The knee joint impedances are estimated based on the internal feedback of the system dynamics, that lead to design the torque compensator to improve the overall control signals in real time. This paper also demonstrates the characteristics of various responses of the system during exercise with human subject. Results have reflected good performances with low position and velocity tracking errors, <span><math><mrow><mo>±</mo><mn>0</mn><mo>.</mo><msup><mn>02</mn><mo>∘</mo></msup></mrow></math></span> and <span><math><mrow><mn>0.04</mn><mi>r</mi><mi>a</mi><mi>d</mi><mo>.</mo><mi>s</mi><mi>e</mi><msup><mi>c</mi><mrow><mo>−</mo><mn>1</mn></mrow></msup></mrow></math></span> during hold phase; and <span><math><mrow><mo>±</mo><mn>0</mn><mo>.</mo><msup><mn>14</mn><mo>∘</mo></msup></mrow></math></span> and <span><math><mrow><mn>0.17</mn><mi>r</mi><mi>a</mi><mi>d</mi><mo>.</mo><mi>s</mi><mi>e</mi><msup><mi>c</mi><mrow><mo>−</mo><mn>1</mn></mrow></msup></mrow></math></span> during motion phse. Though, the limitation of the prototype is its current RoM (limited to <span><math><msup><mn>0</mn><mo>∘</mo></msup></math></span>–<span><math><msup><mn>25</mn><mo>∘</mo></msup></math></span>), the system has potential in the application of RoM exercise for paraplegic or monoplegic patients.</p></div>","PeriodicalId":100288,"journal":{"name":"Cognitive Robotics","volume":"4 ","pages":"Pages 42-61"},"PeriodicalIF":0.0,"publicationDate":"2023-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S266724132300040X/pdfft?md5=e66adda08f021e960ed5946bd42e69d8&pid=1-s2.0-S266724132300040X-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138619946","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Optimizing Speech Emotion Recognition with Hilbert Curve and convolutional neural network","authors":"Zijun Yang ,&nbsp;Shi Zhou ,&nbsp;Lifeng Zhang ,&nbsp;Seiichi Serikawa","doi":"10.1016/j.cogr.2023.12.001","DOIUrl":"10.1016/j.cogr.2023.12.001","url":null,"abstract":"<div><p>In the realm of speech emotion recognition, researchers strive to refine representation methods for improved emotional information capture. Traditional one-dimensional time series classification falls short in expressing intricate emotional patterns present in speech signals, posing challenges in accuracy and robustness. This study introduces an innovative algorithm leveraging Hilbert curves to transform one-dimensional speech data into two-dimensional form, enhancing feature extraction accuracy. A tiling module based on Hilbert curve maximizes Hilbert curve arrangements for improved emotional information capture. Results reveal spatial efficiency gains up to 23,195 times pixel units, enhancing data storage. With an exceptional 98.73% accuracy, the proposed approach traditional methods, affirming its superior emotion classification performance on the same dataset. These empirical findings underscore the effectiveness of our proposed method in advancing speech emotion recognition.</p></div>","PeriodicalId":100288,"journal":{"name":"Cognitive Robotics","volume":"4 ","pages":"Pages 30-41"},"PeriodicalIF":0.0,"publicationDate":"2023-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2667241323000411/pdfft?md5=bfed8ff77493b33cdfb6f93a3ba0a2c9&pid=1-s2.0-S2667241323000411-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138609217","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An improved single short detection method for smart vision-based water garbage cleaning robot","authors":"Anandakumar Haldorai,&nbsp;Babitha Lincy R,&nbsp;Suriya M,&nbsp;Minu Balakrishnan","doi":"10.1016/j.cogr.2023.11.002","DOIUrl":"https://doi.org/10.1016/j.cogr.2023.11.002","url":null,"abstract":"<div><p>These days, plastic trash is exponentially overwhelming our waterways. The catastrophe has attracted global attention at this point. As a result, protecting the environment on the water's surface has received increasing focus. Currently, manpower can be used to clean up contaminated water bodies like ponds, rivers, and oceans. Using the current cleaning approach results in low efficiency and hazard. The detection, collection, sorting, and removal of plastic trash from such water surfaces has been the subject of relatively little robotic research, despite the dire circumstances. From private sources, there are very few individual efforts to be found. In order to attain great efficiency without human assistance or operation, a fully autonomous water surface cleaning robot is proposed in this study. The robot was created to adapt to any type of water body found in the real world. An efficient object identification machine learning technique can be suggested for the creation of autonomous cleaning robots. This study improved the Single Short Detection (SSD) method to recognise objects accurately. Because of the enhanced detection techniques, the robot is able to collect trash on its own. With a mean average precision (mAP) of 94.099 % and a detection speed of up to 64.67 frames per second, experimental findings show that the enhanced SSD has exceptional detection speed and accuracy.</p></div>","PeriodicalId":100288,"journal":{"name":"Cognitive Robotics","volume":"4 ","pages":"Pages 19-29"},"PeriodicalIF":0.0,"publicationDate":"2023-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2667241323000393/pdfft?md5=a8305dcc49d8d37defb2594ad2b10d51&pid=1-s2.0-S2667241323000393-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138738971","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}