2019 International Conference on Machine Learning and Cybernetics (ICMLC)最新文献

{"title":"Phase Retrieval via Wirtinger Flow Algorithm and Its Variants","authors":"Jian-wei Liu, Zhi Cao, Jing Liu, Xiong-lin Luo, Wei-min Li, Nobuyasu Ito, Longteng Guo","doi":"10.1109/ICMLC48188.2019.8949170","DOIUrl":"https://doi.org/10.1109/ICMLC48188.2019.8949170","url":null,"abstract":"Almost three-quarters of the underling information in the light wave field is embodied in the phase. However, the early optical detectors can only record the intensity or amplitude of the light wave field and cannot directly extract the phase information of the light wave field. Therefore, it is necessary to use the measured amplitude or strength to reconstruct the phase information of the object, this problem is denoted phase retrieval. Phase retrieval is a matter of cardinal significance in signal processing and machine learning. The phase retrieval by convex optimization algorithm is ideal but the computational complexity is high. In 2015, Candès proposed a very effective non-convex optimization algorithm-Wirtinger flow algorithm which used spectral initialization to get a better initial value and then gradient iteration to get a promised recovery effect. Subsequently, in line with the idea, a large number of variants are devised, such as: Wirtinger flow(WF), Truncated Wirtinger Flow (TWF), Truncated Amplitude Flow (TAF), Reshaped Wirtinger Flow (RWF), Incremental Truncated Wirtinger Flow (ITWF), Incremental Reshaped Wirtinger Flow (IRWF), Robust Wirtinger Flow (Robust-WF), Sparse Wirtinger Flow (SWF), Median-TWF, Median-RWF, Generalized Wirtinger Flow (GWF), Accelerated Wirtinger Flow (AWF), Thresholded Wirtinger Flow Revisited (THWFR), Thresholded Wirtinger Flow (THWF), Reweighted Wirtinger Flow (REWF), Wirtinger Flow Method With Optimal Stepsize (WFOS), Stochastic Truncated Wirtinger Flow Algorithm (STWF), Stochastic Truncated Amplitude Flow (STAF), Reweighted Amplitude Flow (RAF), Compressive Reweighted Amplitude Flow (CRAF), SPARse Truncated Amplitude flow (SPARTA) and Sparse Wirtinger Flow Algorithm with Optimal Stepsize (SWFOS), etc. This paper analyzes and summarizes these algorithms according to their characteristics such as: initialization method, step size, iteration times, sample complexity, computational complexity, etc., so that readers can intuitively and clearly see the characteristics of each algorithm. Finally, we provide the website of the source code of some algorithms, facilitate to access and use it for readers.","PeriodicalId":221349,"journal":{"name":"2019 International Conference on Machine Learning and Cybernetics (ICMLC)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125432289","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Domain Adaption for Facial Expression Recognition","authors":"Juntong Liu, F. Wu, Wenjin Lu, Bai-Ling Zhang","doi":"10.1109/ICMLC48188.2019.8949178","DOIUrl":"https://doi.org/10.1109/ICMLC48188.2019.8949178","url":null,"abstract":"Facial expression recognition (FER) is a task that recognizes human emotions from their facial expressions. Owing to the lack of large datasets, a FER system is difficult to design, especially for real world environment. In this paper, we propose a new dataset augmentation method for FER and the corresponding training strategy by using similarity preserving generative adversarial network (SPGAN). By borrowing the idea from person re-ID field, we consider dataset augmentation as a domain adaptation task. The SPGAN is first trained on a lab condition dataset and a real world condition dataset to generate domain adapted images, and then CNN models are subsequently trained on the domain adapted images. We test our models on the RAF-DB and SFEW 2.0 datasets to show the improvement when compared it to our baseline. We also report our competitive accuracy when compared it with other state of the art works, which shows promissing results.","PeriodicalId":221349,"journal":{"name":"2019 International Conference on Machine Learning and Cybernetics (ICMLC)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115333366","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Autonomous Cross-Floor Navigation System for a ROS-Based Modular Service Robot","authors":"Wenhui Wang, Yi-Hsing Chien, H. Chiang, Wei-Yen Wang, C. Hsu","doi":"10.1109/ICMLC48188.2019.8949176","DOIUrl":"https://doi.org/10.1109/ICMLC48188.2019.8949176","url":null,"abstract":"In this paper, we present an autonomous cross-floor navigation system including mapping, localization, path planning, and scene recognition based on robot operating system (ROS) architecture. The Gmapping algorithm is utilized to build a 2D map with a laser range-finder, and AMCL algorithm is utilized in the robot localization. Moreover, an improved A* algorithm is proposed to prevent robot from getting too close to the wall. Because our robot needs to navigate in the multi-floor environment, a decision system using deep convolutional neural network (DCNN) is also designed to recognize the current floor and the associated map can be download to the robot system. By training with the scene images of the featured location in each floor, the robot can recognize the current floor and then complete the navigation task. Finally, real test of our robot is conducted to demonstrate the feasibility of the proposed method.","PeriodicalId":221349,"journal":{"name":"2019 International Conference on Machine Learning and Cybernetics (ICMLC)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126679400","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Development of a System to Measure Radioulnar Distance in Wrist-Joint Rotation Using Three-Dimensional Electromagnetic Sensor","authors":"K. Nagamune, Akito Nakano","doi":"10.1109/ICMLC48188.2019.8949284","DOIUrl":"https://doi.org/10.1109/ICMLC48188.2019.8949284","url":null,"abstract":"For sports such as baseball and tennis, there are actions to throw the ball and swing the racket. There are cases injured the wrist joint by repeating this action. One such injury to the wrist joint is triangular fibrocartilage complex (TFCC) injury. TFCC is a part that keeps stability on the ulnar side of the wrist joint scale. So, if the TFCC is injured, the distance between the ulna and the radius will widen due to the wrist rotation, when the injury is severe, pain occurs on the ulnar side of the wrist joint. In the current diagnosis, there is no diagnosis to evaluate the change in distance between the ulna and the radius in the wrist rotation. Therefore, in this study, to quantitatively evaluate the change of distance between the ulna and the radius in TFCC injury, we develop a system to measure the distance between the ulna and the radius in the wrist rotation.","PeriodicalId":221349,"journal":{"name":"2019 International Conference on Machine Learning and Cybernetics (ICMLC)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124966343","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Water Level Prediction at TICH-BUI river in Vietnam Using Support Vector Regression","authors":"Thanh-Tung Nguyen, Hien T. T. Le","doi":"10.1109/ICMLC48188.2019.8949273","DOIUrl":"https://doi.org/10.1109/ICMLC48188.2019.8949273","url":null,"abstract":"In this paper, the support vector regression model is used to predict water levels at a downstream station of the Tich-Bui river basin. The study investigated the effects of rainfall data collected from eight gauging stations and water levels at the downstream station for the performance forecast. The model was set up to forecast water levels at the downstream station before 6-lead-hour, 12-lead-hour, 18-lead-hour and 24-lead-hour. Although the model does not require data on the climate, terrain but the forecast results are accurate. In the case of a water level forecast before 6 hours and 12 hours, the Nash coefficient gives a value of over 98.81% and the RMSE value is less than 0.20 m. This results suggest that the support vector regression model, which the authors use to accurately predict water levels in real time, can be used to warn of floods in Vietnam's rivers.","PeriodicalId":221349,"journal":{"name":"2019 International Conference on Machine Learning and Cybernetics (ICMLC)","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130411081","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"News Recommendation Based on Collaborative Semantic Topic Models and Recommendation Adjustment","authors":"Yu-Shan Liao, Jun-Yi Lu, Duen-Ren Liu","doi":"10.1109/ICMLC48188.2019.8949259","DOIUrl":"https://doi.org/10.1109/ICMLC48188.2019.8949259","url":null,"abstract":"Providing news recommendations is an important trend for online news websites to attract more users and create more benefits. In this research, we propose a novel recommendation approach for recommending news articles. We propose A Collaborative Semantic Topic Model and an ensemble model to predict user preferences based on combining Matrix Factorization with articles' semantic latent topics derived from word embedding and Latent Dirichlet Allocation. The proposed ensemble model is further integrated with a recommendation adjustment mechanism to adjust users' online recommendation lists. We evaluate the proposed approach via offline experiments and online evaluation on a real news website. The experimental result demonstrates that our proposed approach can improve the recommendation quality of recommending news articles.","PeriodicalId":221349,"journal":{"name":"2019 International Conference on Machine Learning and Cybernetics (ICMLC)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132962104","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An Improved Siamese Network for Face Sketch Recognition","authors":"Liang Fan, Han Liu, Y. Hou","doi":"10.1109/ICMLC48188.2019.8949231","DOIUrl":"https://doi.org/10.1109/ICMLC48188.2019.8949231","url":null,"abstract":"Face sketch recognition identifies the face photo from a large face sketch dataset. Some traditional methods are typically used to reduce the modality gap between face photos and sketches and gain excellent recognition rate based on a pseudo image which is synthesized using the corresponded face photo. However, these methods cannot obtain better high recognition rate for all face sketch datasets, because the use of extracted features cannot lead to the elimination of the effect of different modalities' images. The feature representation of the deep convolutional neural networks as a feasible approach for identification involves wider applications than other methods. It is adapted to extract the features which eliminate the difference between face photos and sketches. The recognition rate is high for neural networks constructed by learning optimal local features, even if the input image shows geometric distortions. However, the case of overfitting leads to the unsatisfactory performance of deep learning methods on face sketch recognition tasks. Also, the sketch images are too simple to be used for extracting effective features. This paper aims to increase the matching rate using the Siamese convolution network architecture. The framework is used to extract useful features from each image pair to reduce the modality gap. Moreover, data augmentation is used to avoid overfitting. We explore the performance of three loss functions and compare the similarity between each image pair. The experimental results show that our framework is adequate for a composite sketch dataset. In addition, it reduces the influence of overfitting by using data augmentation and modifying the network structure.","PeriodicalId":221349,"journal":{"name":"2019 International Conference on Machine Learning and Cybernetics (ICMLC)","volume":"69 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133923776","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep License Plate Recognition in Ill-Conditioned Environments With Ill-Conditional Data Augmentation","authors":"C. Lien, Yu-Chun Chien, Fu-Yu Teng, Chih-Chieh Yang","doi":"10.1109/ICMLC48188.2019.8949248","DOIUrl":"https://doi.org/10.1109/ICMLC48188.2019.8949248","url":null,"abstract":"In general, the conventional LPR systems consist of the following modules: feature extraction, license plate locating, character segmentation, and character recognition. The performances of these module are strongly correlated with some low level image features, e.g., edges, colors, and textures. These low level image features can be influenced significantly by the illumination and view angle variations such that the recognition accuracy is degraded. Recently, the deep learning technologies make the conventional vision-based recognition technologies getting significant improvement in terms of feature discrimination and recognition accuracy. In this paper, we aim to develop a novel deep learning based LPR system with the ill-conditional data augmentation. Therefore, this paper is expected to the following contributions. First, we apply the WebGL technology to augment the training database for the ill-conditioned outdoor environments. Second, we apply the YOLOv2 DNN architecture to develop deep license plate recognition system in the ill-conditioned environments with recognition accuracy 98%.","PeriodicalId":221349,"journal":{"name":"2019 International Conference on Machine Learning and Cybernetics (ICMLC)","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132219127","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Predicting Dementia Risk to Depressive Disorder Patients: A classification Approach","authors":"Hsiao-Ting Tseng, Hsiao-Chi Li, Chia-Lun Lo, Tai-Hsiang Shen, Shu-Chiung Lin","doi":"10.1109/ICMLC48188.2019.8949191","DOIUrl":"https://doi.org/10.1109/ICMLC48188.2019.8949191","url":null,"abstract":"The WHO identified depressive disorder as one of the three major diseases in the 21st century and studies have shown that patients with depression are more likely than nondepression to have dementia in the future. However, although there are many related studies that point out that depressive disorder is one of the important factor of dementia, however, these findings are not consistent. In addition, there has been no study of evidence-based construction of dementia prediction model of depressive disorder patients for clinical practice. Therefore, this study will use supervised learning techniques to construct a follow-up dementia prediction model for depressive disorder patients to assist depressive disorder patients and their medical staffs to predict his/her possible risk of suffering from dementia, and then develop early intervention and prevention measures.","PeriodicalId":221349,"journal":{"name":"2019 International Conference on Machine Learning and Cybernetics (ICMLC)","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133749987","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Single-Image Super-Resolution via Multiple Matrix-Valued Kernel Regression","authors":"Yi Tang, Zuo Jiang, Junhua Chen","doi":"10.1109/ICMLC48188.2019.8949261","DOIUrl":"https://doi.org/10.1109/ICMLC48188.2019.8949261","url":null,"abstract":"Single-image super-resolution focuses on learning a mapping to recover high-resolution images from given low-resolution images with the help of a set of paired images. Matrix-valued operators serve as an efficient mapping to super-resolve low-resolution images. However, most existed matrix-valued based super-resolution algorithms limit matrix-valued operators as linear mappings. Multiple matrix-valued operators based algorithm is introduced for improving the performance of matrix-value operators in single-image super-resolution. Taking advantages of the non-linear style of multiple matrix-valued operators, we have more accurate super-resolved images. The experimental results show the efficiency and effectiveness of the reported multiple matrix-valued operator learning based super-resolution algorithm.","PeriodicalId":221349,"journal":{"name":"2019 International Conference on Machine Learning and Cybernetics (ICMLC)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133717505","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}