A5077550371- Advanced Neural Network Applications
- Advanced Image and Video Retrieval Techniques
- Magnetic Bearings and Levitation Dynamics
- Electric Motor Design and Analysis
- Robotics and Sensor-Based Localization
- Remote-Sensing Image Classification
- Image Enhancement Techniques
- Multimodal Machine Learning Applications
- Advanced Data Compression Techniques
- Water Quality Monitoring Technologies
- Advanced Sensor and Control Systems
- Sensorless Control of Electric Motors
- Advanced Battery Technologies Research
- Power Systems and Renewable Energy
- Advanced Image Fusion Techniques
- Domain Adaptation and Few-Shot Learning
- Image and Signal Denoising Methods
- Machine Learning and Data Classification
Nanjing University of Aeronautics and Astronautics
2024-2025
Shenyang University of Chemical Technology
2025
Chinese Academy of Sciences
2025
Shenyang Institute of Automation
2025
State Key Laboratory of Information Engineering in Surveying Mapping and Remote Sensing
2019-2022
Wuhan University
2019-2022
Dalian University of Technology
2021
To boost the object grabbing capability of underwater robots for open-sea farming, we propose a new dataset (UDD) consisting three categories (seacucumber, seaurchin, and scallop) with 2,227 images. best our knowledge, it is first 4K HD collected in real farm. We also novel Poisson-blending Generative Adversarial Network (Poisson GAN) an efficient detection network (AquaNet) to address two common issues within related datasets: class-imbalance problem mass small object, respectively....
Owing to the rapid development of deep neural networks, prominent advances have been recently achieved in semantic segmentation remote sensing images. As vital components computer vision, segmentation, and edge detection strong correlation whether extracted features or task objective. Prior studies treated as a postprocessing operation they implicitly combined two tasks. We consider that pixels around edges are easy be misdivided because prevalence intraclass inconsistencies interclass...
The lithium-ion battery is increasingly critical in the fields of electric vehicles and sustainable energy. Accurate prediction Remaining Useful Life (RUL) batteries essential to mitigate risks minimize potential losses. This paper introduces a novel fusion algorithm, FPSOGWO, which combines Particle Swarm Optimization (PSO) Grey Wolf (GWO) with an enhanced Circle chaos strategy random adaptive weighting strategy. FPSOGWO integrated Support Vector Regression (FPSOGWO-SVR), was utilized...
Compressing hyperspectral images (HSIs) into compact representations under the premise of ensuring high-quality reconstruction is an essential task in HSI processing. However, existing compression methods usually encode by smoothing due to low-frequency information occupying a prominent component most images. Consequently, these fail capture sufficient structural information, especially low bit rates, often causing inferior reconstruction. To address this problem, we propose here edge-guided...
Remarkable improvements have been seen in the semantic segmentation of remote-sensing images. As an effective structure to aggregate shallow information and deep information, encoder–decoder has widely used many state-of-the-art models, but it possesses two drawbacks that not fully addressed. On one hand, fuses features obtained from layers directly; despite harvesting some detailed also brings noisy owing poor discriminant ability layers. other existing merely high-level generated by last...
Global context information is vital in visual understanding problems, especially pixel-level semantic segmentation. The mainstream methods adopt the self-attention mechanism to model global information. However, pixels belonging different classes usually have weak feature correlation. Modeling correlation matrix indiscriminately extremely redundant mechanism. In order solve above problem, we propose a hierarchical network differentially homogeneous with strong correlations and heterogeneous...
Semantic segmentation in high resolution aerial image is faced with a challenge caused by ubiquitous fine-structure objects. Traditional encoder-decoder structure losses some detail information during the process of down-sampling, which harmful to location In this work, we present multi-resolution dense encoder and stack decoder network deal problem. On one hand, embeds shallow detailed feature into deep semantic through proposed information-reserved down-sampling method called CE-Pooling....
Global context information is vital in visual understanding problems, especially pixel-level semantic segmentation. The mainstream methods adopt the self-attention mechanism to model global information. However, pixels belonging different classes usually have weak feature correlation. Modeling correlation matrix indiscriminately extremely redundant mechanism. In order solve above problem, we propose a hierarchical network differentially homogeneous with strong correlations and heterogeneous...