A5100366838- Infrared Target Detection Methodologies
- Calibration and Measurement Techniques
- Advanced Image Fusion Techniques
- Advanced Measurement and Detection Methods
- Quantum Computing Algorithms and Architecture
- Music and Audio Processing
- Optical Network Technologies
- Quantum Information and Cryptography
- Advanced Chemical Sensor Technologies
- Image and Signal Denoising Methods
- Optical Systems and Laser Technology
- Interactive and Immersive Displays
- Neural Networks and Reservoir Computing
- Meat and Animal Product Quality
- Advanced Vision and Imaging
- Speech and Audio Processing
- Virtual Reality Applications and Impacts
- Human Pose and Action Recognition
- Radio Astronomy Observations and Technology
- Quantum-Dot Cellular Automata
- Machine Fault Diagnosis Techniques
- Fractal and DNA sequence analysis
- Optical measurement and interference techniques
- Digital Filter Design and Implementation
- Animal Behavior and Welfare Studies
Wuhan University
2015-2025
Southwestern Institute of Physics
2023
Technical University of Munich
2023
Xi’an University of Posts and Telecommunications
2022
LMU Klinikum
2020
Ludwig-Maximilians-Universität München
2020
Huazhong University of Science and Technology
2013-2014
Anhui University of Science and Technology
2008
Melody preservation is crucial in singing voice conversion (SVC). However, many scenarios, audio often accompanied with background music (BGM), which can cause distortion and interfere the extraction of melody other key features, significantly degrading SVC performance. Previous methods have attempted to address this by using more robust neural network-based extractors, but their performance drops sharply presence complex accompaniment. Other approaches involve performing source separation...
Exploiting deep learning techniques to automatically analyze multi-spectral remote sensing imagery plays an essential role in urban land cover and use classification. However, the computation power required large earth observation data with complex machine models for this task becomes intractable bottleneck. Leveraging quantum computing might tackle challenge. In paper, we present two hybrid quantum-classical frameworks. They both exploit extract features from images efficiently classical...
Fault diagnosis of rotating machinery plays a significant role in the reliability and safety modern industrial systems, which generally requires collaborative fault by features extracted from multiple sensors since multi-channel vibration signals carry wealth information. However, there is remaining obstacle for multi-source monitoring data: integration multisensory data. Hence, novel framework proposed First, composite multivariate multi-scale symbolic dynamic entropy to extract features....
We investigate the use of a handheld spherical object as controller for locomotion in VR. Rotating controls avatar movement two different ways: As zero order controller, it is continuously rotated to target position if rolling ball on floor. first tilted like joystick determine direction and speed movement. describe how our prototype was built from low-cost commercially available hardware discuss design decisions. Then we evaluate both techniques user study (N=20) compare them established...
Raw infrared images suffer the problem of low contrast. In this paper, a novel saliency weight based image enhancement method is proposed. For purpose preserving relative thermal distribution information, global-mapping preferred in work. To address under-enhancement on object areas probability-histogram methods, spatial information utilized to develop new intensity levels for take place occurrence probability, named weight. weight, values objects-related are larger than background-related...
The infrared focal plane array (IRFPA) suffers from the problem of non-uniform photoelectric responsivity detectors, namely nonuniformity. This paper aims at removing fixed pattern noise (FPN) brought by nonuniformity in IRFPA through scene-based correction (SBNUC). Different traditional neural network-based methods where spatial filters are adopted, a new approach acquiring desired image (the target optimization task) is proposed, which based on sparse representation theory. corrected...
Variational Autoencoders (VAEs) constitute a crucial component of neural symbolic music generation, among which some works have yielded outstanding results and attracted considerable attention. Nevertheless, previous VAEs still encounter issues with overly long feature sequences generated lack contextual coherence, thus the challenge modeling multi-track remains unaddressed. To this end, we propose Multi-view MidiVAE, as one pioneers in VAE methods that effectively model generate music. The...
Continuous-variable quantum key distribution (CV-QKD) offers an approach to achieve a potential high secret rate (SKR) in metropolitan areas. There are several challenges developing practical CV-QKD system from the laboratory real world. One of most significant points is that it really hard adapt different optical fiber conditions for systems with unified hardware. Thus, how improve performance field without modification hardware very important. Here, systematic optimization method,...