Image classification has always been a hot research direction in the world, and emergence of deep learning promoted development this field. Convolutional neural networks (CNNs) have gradually become mainstream algorithm for image since 2012, CNN architecture applied to other visual recognition tasks (such as object detection, localization, semantic segmentation) is generally derived from network classification. In wake these successes, CNN-based methods emerged remote sensing scene achieved...

This paper focuses on bearing fault diagnosis with limited training data. A major challenge in is the infeasibility of obtaining sufficient samples for every type under all working conditions. Recently deep learning based methods have achieved promising results. However, most these require large amount In this study, we propose a neural network few-shot approach rolling Our model siamese network, which learns by exploiting sample pairs same or different categories. Experimental results over...

Abstract A major challenge in materials design is how to efficiently search the vast chemical space find with desired properties. One effective strategy develop sampling algorithms that can exploit both explicit knowledge and implicit composition rules embodied large database. Here, we propose a generative machine learning model (MatGAN) based on adversarial network (GAN) for efficient generation of new hypothetical inorganic materials. Trained from ICSD database, our GAN generate not...

Intelligent machine health monitoring and fault diagnosis are becoming increasingly important for modern manufacturing industries. Current approaches mostly depend on expert-designed features building prediction models. In this paper, we proposed IDSCNN, a novel bearing algorithm based ensemble deep convolutional neural networks an improved Dempster-Shafer theory evidence fusion. The take the root mean square (RMS) maps from FFT (Fast Fourier Transformation) of vibration signals two sensors...

Prognostics, such as remaining useful life (RUL) prediction, is a crucial task in condition-based maintenance. A major challenge data-driven prognostics the difficulty of obtaining sufficient number samples failure progression. However, for traditional machine learning methods and deep neural networks, enough training data prerequisite to train good prediction models. In this work, we proposed transfer algorithm based on Bi-directional Long Short-Term Memory (BLSTM) recurrent networks RUL...

We adopted actual intelligent production requirements and proposed a tiny part defect detection method to obtain stable accurate real-time system solve the problems of manually setting conveyor speed industrial camera parameters in for factory products. First, we considered important influences properties parts environmental on its stability. Second, established correlation model between capability coefficient moving conveyor. Third, algorithm that are based single short detector network...

Convolutional neural networks (CNNs) with log-mel audio representation and CNN-based end-to-end learning have both been used for environmental event sound recognition (ESC). However, features can be complemented by learned from the raw waveform an effective fusion method. In this paper, we first propose a novel stacked CNN model multiple convolutional layers of decreasing filter sizes to improve performance models either feature input or input. These two are then combined using...

With the rapid development of machine learning, its powerful function in vision field is increasingly reflected. The combination and robotics to achieve same precise fast grasping as that humans requires high-precision target detection recognition, location reasonable grasp strategy generation, which ultimate goal global researchers one prerequisites for large-scale application robots. Traditional learning has a long history good achievements image processing robot control. CNN...

Currently gear fault diagnosis is mainly based on vibration signals with a few studies acoustic signal analysis. However, acquisition limited by its contact measuring while traditional acoustic-based relies heavily prior knowledge of processing techniques and diagnostic expertise. In this paper, novel deep learning-based method proposed sound By establishing an end-to-end convolutional neural network (CNN), the time frequency domain can be fed into model as raw without feature engineering....

Traffic congestion prediction is critical for implementing intelligent transportation systems improving the efficiency and capacity of networks. However, despite its importance, traffic severely less investigated compared to flow prediction, which partially due severe lack large-scale high-quality data advanced algorithms. This paper proposes an accessible general workflow acquire create datasets based on image analysis. With this we a dataset named Seattle Area Congestion Status (SATCS) map...

Many text mining tasks such as retrieval, summarization, and comparisons depend on the extraction of representative keywords from main text. Most existing keyword algorithms are based discrete bag-of-words type word representation In this paper, we propose a patent algorithm (PKEA) distributed Skip-gram model for classification. We also develop set quantitative performance measures evaluation information gain cross-validation, Support Vector Machine (SVM) classification, which valuable when...

Machine Reading Comprehension (MRC) is a challenging Natural Language Processing (NLP) research field with wide real-world applications. The great progress of this in recent years mainly due to the emergence large-scale datasets and deep learning. At present, lot MRC models have already surpassed human performance on various benchmark despite obvious giant gap between existing genuine human-level reading comprehension. This shows need for improving datasets, evaluation metrics, move current...

In the industrial field, anthropomorphism of grasping robots is trend future development, however, basic vision technology adopted by robot at this stage has problems such as inaccurate positioning and low recognition efficiency. Based on practical problem, in order to achieve more accurate objects, an object detection method for based improved YOLOv5 was proposed paper. Firstly, platform designed, wooden block image data set being proposed. Secondly, Eye-In-Hand calibration used obtain...

As living standards improve, modern products need to meet increasingly diversified and personalized user requirements. Traditional product design methods fall short due their strong subjectivity, limited survey scope, lack of real-time data, poor visual display. However, recent progress in big data artificial intelligence (AI) are bringing a transformative AI-driven methodology with significant impact on many industries. Big the lifecycle contains valuable information, such as customer...