A5102828980- Advanced Neural Network Applications
- Video Surveillance and Tracking Methods
- Gear and Bearing Dynamics Analysis
- Machine Fault Diagnosis Techniques
- Industrial Vision Systems and Defect Detection
- Advanced Image and Video Retrieval Techniques
- Remote-Sensing Image Classification
- Educational Reforms and Innovations
- Topic Modeling
- Advanced Decision-Making Techniques
- Engineering Diagnostics and Reliability
- Acoustic Wave Phenomena Research
- Image Enhancement Techniques
- Robotics and Sensor-Based Localization
- Infrastructure Maintenance and Monitoring
- Noise Effects and Management
- Fault Detection and Control Systems
- Visual Attention and Saliency Detection
- Image Processing Techniques and Applications
- Evaluation and Optimization Models
- Vehicle Noise and Vibration Control
- Image and Object Detection Techniques
- Brain Tumor Detection and Classification
- Digital Media Forensic Detection
- MRI in cancer diagnosis
PLA Army Engineering University
2019-2025
Traditional fault diagnosis methods often require extracting features from raw vibration signals based on prior knowledge, which are then input into intelligent classifiers for pattern recognition. This process is prone to information loss and can be inaccurate when relying human experience identification. To address this issue, paper proposes an classification model rolling bearings Fast Fourier Transform (FFT) combined with a time convolutional network (SE-TCN) incorporating attention...
Traffic sign detection systems provide important road control information for unmanned driving or auxiliary driving. In this paper, the Faster region with a convolutional neural network (R-CNN) traffic in real situations has been systematically improved. First, first step proposal algorithm based on simplified Gabor wavelets (SGWs) and maximally stable extremal regions (MSERs) is proposed. way, priori obtained will be used improving R-CNN. This part of our method named as highly possible...
This paper addresses the challenge of small object detection in remote sensing image recognition by proposing an improved YOLOv8-based lightweight attention cross-scale feature fusion model named LACF-YOLO. Prior to backbone network outputting maps, this introduces a module, Triplet Attention, and replaces Concatenation with Fusion (C2f) more convenient higher-performing dilated inverted convolution layer acquire richer contextual information during extraction phase. Additionally, it employs...
Effective intelligent fault diagnosis of bearings is important for improving safety and reliability machine. Benefiting from the training advantages, deep learning method can automatically adaptively learn more abstract high-level features without much priori knowledge. To realize representative mining automatic recognition bearing health condition, a diagnostic model stacked sparse denoising autoencoder (SSDAE) which combines (SAE) (DAE) proposed in this paper. The criterion SAE, corrupting...
Multispectral pedestrian detection based on deep learning can provide a robust and accurate under different illumination conditions, which has important research significance in safety. In order to reduce the log-average miss rate of object new one-stage detector suitable for multispectral is proposed. First, realize complementarity between information flows two modalities feature extraction stage loss, low-cost cross-modality complementary module (CFCM) Second, suppress background noise...
Abstract MRI (magnetic resonance imaging) images can effectively show the placental haemorrhage area. In view of special properties and real‐time detection requirements images, this paper has systematically improved single‐shot multi‐box detector (SSD) target algorithm (M‐SSD). First, taking advantage particularity image, maximum stable extremum region (MSER) was used as anchor proposal network which integrated information into feature layer SSD to avoid hungry traversal original algorithm....
Bearing fault is a process of gradual development and deepening. In the early stage fault, if it can be found out in time taken reasonable prevention elimination measures, we avoid serious losses safety accidents. Therefore, feature extraction analysis weak has important practical significance. this paper, an improved multiscale permutation entropy (IMPE) method was proposed to overcome shortcomings coarse-grained process. order solve problem that only considering single sequence under...
Object detection in remote sensing imagery is a challenging task the field of computer vision and has high research value. To improve classification accuracy positioning object detection, we propose new multi-scale oriented detector suitable for small objects. Firstly, feature fusion network based on information balance (IBFF) proposed to reduce reuse different layers' features from backbone interference redundant premise that output have sufficient information, retain enough shallow detail...
Italy is a major manufacturing exporter. The export revenue takes up significant proportion of its national income to which exported commodities Italian brands contribute great deal. This why the authority has been making continuous efforts crack down on intellectual property infringement for protection exports and ultimately stability economy. Once entering Chinese market where counterfeit consumer goods go rampant, enterprises have urge fight against different kinds counterfeiting under...
With the development of unmanned vehicles and other technologies, technical demand for scene semantic segmentation is more intense. Semantic requires not only rich high-level information, but also detail information to ensure accuracy task. Using a multipath structure process underlying can improve efficiency while ensuring accuracy. In order some small thin objects, guided multilateral network proposed. Firstly, in model efficiency, trilateral parallel designed, including context fusion...