Automatic identification and quantification of dense microcracks in high-performance fiber-reinforced cementitious composites through deep learning-based computer vision
شناسایی و تعیین کمی ترکهای متراکم در کامپوزیت های سیمانی با عملکرد بالا با استفاده از دید رایانه ای مبتنی بر یادگیری عمیق-2021
High-performance fiber-reinforced cementitious composites (HPFRCCs) feature high mechanical strengths, crack resistance, and durability. Under excessive loading, HPFRCCs demonstrate dense microcracks that are difficult to identify using existing methods. This study presents a computer vision method for identification, quantification, and visualization of microcracks in HPFRCCs based on deep learning. The presented method integrates multiple deep learning models and computer vision techniques in a hierarchical architecture. The crack pattern (e.g., number, width, and spacing of cracks) are automatically determined from pictures without human intervention. This study shows that the presented method achieves an accuracy of 0.992 for crack detection and an accuracy finer than 50 μm (R2 > 0.984) for quantification of crack width when deep learning models are trained using only 200 pictures of HPFRCCs and 200 pictures of conventional concrete with incorporation of data augmentation. The presented method is expected to be also applicable to other materials featuring complex cracks.
Keywords: Computer vision | Crack detection | Crack quantification | Deep learning | High-performance fiber reinforced | cementitious composites (HPFRCC) | Microcrack
Dynamic 3D image simulation of basketball movement based on embedded system and computer vision
شبیه سازی تصویر پویا سه بعدی حرکت بسکتبال بر اساس سیستم تعبیه شده و بینایی ماشین-2021
Traditional empirical basketball teaching methods can be repeated, affecting serious basketball training efficiency and the acquisition of technical essentials. Based on this problem, the basketball training reproduction framework is built utilizing augmented reality innovation. The framework sets up a virtual reenactment model ofa ballplayer planning a player’s track. Simultaneously, as a helping player, it captures the basketball player’sactual situation, compares them with the simulated trajectories, and provides more targeted training. Based on virtual reality-based Virtual Data Augmentation Technology (VDRT), basketball technology’s teaching mode allows players to acquire key points of sports skills and significantly improve basketball players’ training efficiency as soon as possible. With the quick improvement of current science and innovation, for example, center, science and innovation, and electronic data innovation, more educational activities are being applied. However, modern educational methods’ important content is to master and use modern educational equipment and processes. This article uses basic concepts, characteristics, and virtual reality techniques and literature and information methods to explain the types of role play in basketball lessons. Finally, it analyzes the application programs of basketball theory education, technical education, tactical instruction and educational competitions that provide scientific standards for future basketball education reform.
Keywords: Basketball movement | virtual data reinforcement technique (VDRT) | Field Programmable Gate Array (FPGA)
Zero shot augmentation learning in internet of biometric things for health signal processing
یادگیری تقویتی صفر در اینترنت اشیا بیومتریک برای پردازش سیگنال سلامتی-2021
In recent years, the number of Internet of Things (IoT) devices has increased rapidly. The Internet of Biometric Things (IoBT) can process biometrics and health signals, and it will greatly extend the range of biometric applications. The analysis of health signals in the IoBT can use computer-aided diagnosis techniques. However, most of the existing computer-aided diagnosis methods are developed for common diseases and are not suitable for rare diseases. Zero shot learning is a potential method for the computer- aided diagnosis of rare diseases because it can identify objects of unknown categories. However, the ex- isting zero shot learning methods are based on attribute learning and rely on an attribute dataset. There is no attribute dataset for health signal processing. Therefore, the existing zero shot learning methods are not suitable for health signal processing. Based on the above background, we propose a zero shot aug- mentation learning model (ZSAL) in the IoBT for health signal processing. First, an expert doctor identiﬁes the contour of a lesion and selects a background image without a lesion. Second, the computer automatically generates virtual images using zero shot augmentation technology. Finally, the generated virtual dataset is used to train a convolutional classiﬁer, and then we apply the classiﬁer to the computer-aided diagnosis of actual medical images. The experiment shows the eﬃciency and effectiveness of our method.© 2021 Elsevier B.V. All rights reserved.
Keywords: Internet of biometric things | Zero shot learning | Data augmentation | Health signal processing
On the channel density of EEG signals for reliable biometric recognition
چگالی کانال سیگنال های EEG برای تشخیص بیومتریک قابل اعتماد-2021
Electroencephalography (EEG) provides appealing biometrics by encompassing unique attributes including robustness against forgery, privacy compliance, and aliveness detection. Among the main challenges in deploying EEG biometric systems in real-world applications, stability and usability are two important ones. They respectively reflect the capacity of the system to provide stable performance within and across different states, and the ease of use of the system. Previous studies indicate that the usability of an EEG biometric system is largely affected by the number of electrodes and reducing channel density is an effective way to enhance usability. However, it is still unclear what is the impact of channel density on recognition performance and stability. This study examines this issue for systems using different feature extraction and classification methods. Our results reveal a trade-off between channel density and stability. With low-density EEG, the recognition accuracy and stability are compromised to varying degrees. Based on the analysis, we propose a framework that integrates channel density augmentation, functional connectivity estimation and deep learning models for practical and stable EEG biometric systems. The framework helps to improve the stability of EEG biometric systems that use consumer-grade low channel density devices, while retaining the advantages of high usability.
Keywords: EEG biometrics | Data augmentation | Deep learning | Current source density
Color Image Enhancement based on Gamma Encoding and Histogram Equalization
بهبود تصویر رنگی بر اساس رمزگذاری گاما و یکسان سازی هیستوگرام-2021
Image Enhancement is used as a preprocessing step in many computer vision applications. It provides enhanced input for other computerized image processing methods. Many preprocessing techniques can be applied to images depending on the application domain. In this paper we are proposing an image enhancement technique for color images that can be used as preprocessing step in many computer vision applications. It can also be used as a data augmentation technique in object detection. Luminance component of images is sometimes not captured by cameras and displayed by monitors properly. To remove this drawback of devices we have used gamma encoding. Four different values of gamma are evaluated depending on the quality of images. Image is then converted into YUV Color space. Y component represents the luminance. U and V components represent color. After that Contrast Limited Adaptive Histogram Equalization is applied to the Y component to improve the contrast of the image. The results are compared with the state-of-the-art methods on the basis of Peak Signal to noise Ratio (PSNR) and Mean Square Error (MSE). Quantitative results show that proposed algorithm results in improved value of PSNR and decreased value of MSE as compared to existing methods. Qualitative comparison is also done and results show improvement over the existing techniques.© 2020 Elsevier Ltd. All rights reserved. Selection and peer-review under responsibility of the scientific committee of the International Conference on Materials, Manufacturing and Mechanical Engineering for Sustainable Developments-2020.
Keywords: Histogram | Intensity | Luminance | Contrast stretching
Transform domain representation-driven convolutional neural networks for skin lesion segmentation
انتقال شبکه های عصبی کانولوشن نمایندگی محور دامنه برای تقسیم بندی ضایعه پوستی-2020
Automated diagnosis systems provide a huge improvement in early detection of skin cancer, and con- sequently, contribute to successful treatment. Recent research on convolutional neural network has achieved enormous success in segmentation and object detection tasks. However, these networks require large amount of data that is a big challenge in medical domain where often have insufficient data and even a pretrained model on medical images can be hardly found. Lesion segmentation as the initial step of skin cancer analysis remains a challenging issue since datasets are small and include a variety of im- ages in terms of light, color, scale, and marks which have led researchers to use extensive augmentation and preprocessing techniques or fine tuning the network with a pretrained model on irrelevant images. A segmentation model based on convolutional neural networks is proposed in this study for the tasks of skin lesion segmentation and dermoscopic feature segmentation. The network is trained from scratch and despite the small size of datasets neither excessive data augmentation nor any preprocessing to remove artifacts or enhance the images are applied. Alternatively, we investigated incorporating image represen- tations of the transform domain to the convolutional neural network and compared to a model with more convolutional layers that resulted in 6% higher Jaccard index and has shorter training time. The model improved by applying CIELAB color space and the performance of the final proposed architecture is evaluated on publicly available datasets from ISBI challenges in 2016 and 2017. The proposed model has resulted in an improvement of as much as 7% for the segmentation metrics and 17% for the fea- ture segmentation, which demonstrates the robustness of this unique hybrid framework and its future applications as well as further improvement.
Keywords: Convolutional neural network | Dermoscopic features | Melanoma | Skin lesion segmentation | Transform domain
Semi-supervised gear fault diagnosis using raw vibration signal based on deep learning
تشخیص خطای دنده نیمه نظارت شده با استفاده از سیگنال لرزش خام بر اساس یادگیری عمیق-2019
In aerospace industry, gears are the most common parts of a mechanical transmission system. Gear pitting faults could cause the transmission system to crash and give rise to safety disaster. It is always a challenging problem to diagnose the gear pitting condition directly through the raw signal of vibration. In this paper, a novel method named augmented deep sparse autoencoder (ADSAE) is proposed. The method can be used to diagnose the gear pitting fault with relatively few raw vibration signal data. This method is mainly based on the theory of pitting fault diagnosis and creatively combines with both data augmentation ideology and the deep sparse autoencoder algorithm for the fault diagnosis of gear wear. The effectiveness of the proposed method is validated by experiments of six types of gear pitting conditions. The results show that the ADSAE method can effectively increase the network generalization ability and robustness with very high accuracy. This method can effectively diagnose different gear pitting conditions and show the obvious trend according to the severity of gear wear faults. The results obtained by the ADSAE method proposed in this paper are compared with those obtained by other common deep learning methods. This paper provides an important insight into the field of gear fault diagnosis based on deep learning and has a potential practical application value.
KEYWORDS : Deep learning | Gear pitting diagnosis | Gear teeth | Raw vibration signal | Semi-supervised learning | Sparse autoencoder
از تشخیص موسیقی نوری تا تشخیص موسیقی دستنویس: یک مبنا
سال انتشار: 2019 - تعداد صفحات فایل pdf انگلیسی: 8 - تعداد صفحات فایل doc فارسی: 24
تشخیص موسیقی نوری (OMR) شاخه ای از تجزیه و تحلیل تصویری سند محسوب می شود که در پی تبدیل تصاویر پارتیتورها به صورتی قابل خوانش توسط کامپیوتر می باشد. به رغم دهه ها تحقیق، تشخیص پارتیتورهای دستنویس که در اصل نتنگاری غربی است، همچنان یک مساله مفتوح بوده و آثار معدودی وجود دارند که تنها بر روی مرحله خاصی از OMR تمرکز نموده اند. در اثر حاضر، ما سیستم کاملی از تشخیص موسیقی دستنویس (HMR) را بر اساس شبکه های عصبی بازگشتی پیچشی، داده افزایی و یادگیری انتقالی پیشنهاد نمودیم که می تواند به عنوان مبنایی برای جامعه تحقیقاتی عمل نماید.
کلمات کلیدی: تشخیص موسیقی نوری | تشخیص موسیقی دست نویس | تجزیه و تحلیل تصویر و شناسایی تصویر | شبکه های عصبی عمیق | LSTM
|مقاله ترجمه شده|
Deep learning for waveform identification of resting needle electromyography signals
یادگیری عمیق برای شناسایی شکل موج سیگنالهای الکترومیوگرافی سوزن ساکن-2019
Objective: Given the recent advent in machine learning and artificial intelligence on medical data analysis, we hypothesized that the deep learning algorithm can classify resting needle electromyography (n- EMG) discharges. Methods: Six clinically observed resting n-EMG signals were used as a dataset. The data were converted to Mel-spectrogram. Data augmentation was then applied to the training data. Deep learning algorithms were applied to assess the accuracies of correct classification, with or without the use of pre-trained weights for deep-learning networks. Results: While the original data yielded the accuracy up to 0.86 on the test dataset, data-augmentation up to 200,000 training images showed significant increase in the accuracy to 1.0. The use of pre-trained weights (fine tuning) showed greater accuracy than ‘‘training from scratch”. Conclusions: Resting n-EMG signals were successfully classified by deep-learning algorithm, especially with the use of data augmentation and transfer learning techniques. Significance: Computer-aided signal identification of clinical n-EMG testing might be possible by deeplearning algorithms.
Keywords: Needle electromyography | Deep learning | Artificial neural network | Data augmentation | Resting discharge
Application of deep transfer learning for automated brain abnormality classification using MR images
کاربرد یادگیری انتقال عمیق برای طبقه بندی خودکار ناهنجاری مغزی با استفاده از تصاویر MR-2019
Magnetic resonance imaging (MRI) is the most common imaging technique used to detect abnormal brain tumors. Traditionally, MRI images are analyzed manually by radiologists to detect the abnormal conditions in the brain. Manual interpretation of huge volume of images is time consuming and difficult. Hence, computer-based detection helps in accurate and fast diagnosis. In this study, we proposed an approach that uses deep transfer learning to automatically classify normal and abnormal brain MR images. Convolutional neural network (CNN) based ResNet34 model is used as a deep learning model. We have used current deep learning techniques such as data augmentation, optimal learning rate finder and fine-tuning to train the model. The proposed model achieved 5-fold classification accuracy of 100% on 613 MR images. Our developed system is ready to test on huge database and can assist the radiologists in their daily screening of MR images.
Keywords: MRI classification | Abnormal brain images | Deep transfer learning | CNN