A5101513129- Chaos-based Image/Signal Encryption
- Advanced Steganography and Watermarking Techniques
- Digital Media Forensic Detection
- Quantum Computing Algorithms and Architecture
- Speech and Audio Processing
- Indoor and Outdoor Localization Technologies
- Mathematical Dynamics and Fractals
- stochastic dynamics and bifurcation
- Nonlinear Dynamics and Pattern Formation
- Adversarial Robustness in Machine Learning
- Mechanical and Optical Resonators
- Ultra-Wideband Communications Technology
- Geotechnical Engineering and Soil Mechanics
- Quantum Mechanics and Applications
- Landslides and related hazards
- Machine Learning in Materials Science
- Generative Adversarial Networks and Image Synthesis
- Fractal and DNA sequence analysis
- DNA and Biological Computing
- Granular flow and fluidized beds
- Quantum Information and Cryptography
- Nuclear Physics and Applications
University of Electronic Science and Technology of China
2019-2023
Joint Center for Quantum Information and Computer Science
2019
University of California, Los Angeles
2016
This paper proposes a high-quality color image compression-encryption scheme based on chaos and block permutation. In this scheme, the digital is first converted sampled in YCbCr gamut, then coefficients sub-blocks are extracted for compression coding after 8×8 post-blocking discrete cosine transformation (DCT) to frequency domain. Then, an encryption algorithm using chaos-based permutation two-round row-column diffusion designed compressed domain information. terms of security, module...
This paper proposes a dynamic RNA-encoded color image encryption scheme based on chain feedback structure. Firstly, the pure is decomposed into red, green, and blue components, then chaotic sequence plaintext association introduced to encrypt red component. Secondly, intermediate ciphertext obtained by diffusion after bit-level permutation, RNA encoding, operation rules, decoding. Finally, enhance security of cryptosystem, green components are repeatedly encrypted using mechanism associated...
The study of quantum generative models is well-motivated, not only because its importance in machine learning and chemistry but also the perspective implementation on near-term machines. Inspired by previous studies adversarial training classical models, we propose first design Wasserstein Generative Adversarial Networks (WGANs), which has been shown to improve robustness scalability even noisy hardware. Specifically, a definition semimetric between data, inherits few key theoretical merits...
With the arrival of age big data, amount and types data in process information transmission have increased significantly, full-disk encryption mode used by traditional algorithms has certain limitations times. In order to further improve bandwidth efficiency digital images effectiveness image transmission, this paper proposes an algorithm high-quality restoration using DCT frequency-domain compression coding chaos. Firstly, hash value is for generation key with plaintext correlation, then...
In order to further improve the information effectiveness of digital image transmission, an image-encryption algorithm based on 2D-Logistic-adjusted-Sine map (2D-LASM) and Discrete Wavelet Transform (DWT) is proposed. First, a dynamic key with plaintext correlation generated using Message-Digest Algorithm 5 (MD5), 2D-LASM chaos obtain chaotic pseudo-random sequence. Secondly, we perform DWT from time domain frequency decompose low-frequency (LF) coefficient high-frequency (HF) coefficient....
Machine learning offers a powerful framework for validating and predicting atomic mass. We compare three improved neural network methods representation extrapolation mass prediction. The method, adopting macroscopic-microscopic approach treating complex nuclear effects as output labels, achieves superior accuracy in AME2020, yielding much lower root-mean-square deviation of 0.122 MeV the test set, significantly than alternative methods. It also exhibits better performance when AME2020 from...
As an important resource to realize quantum information, correlation displays different behaviors, freezing phenomenon and non-localization, which are dissimilar the entanglement classical correlation, respectively. In our setup, ordering of is represented for quantization methods by considering open system scenario. The machine learning method (neural network method) then adopted train construction a bridge between R\`{e}nyi discord ($\alpha=2$) geometric (Bures distance) $X$ form states....
The simulation of a large number particles requires unacceptable computational time that is the most critical problem existing in industrial application DEM. Coarse graining promising approach to facilitate DEM problems. While current coarse framework often developed an ad-hoc manner, leading different formulations and solution accuracy efficiency. Therefore, this paper, techniques have been carefully analysed by exact scaling law which can provide theory basis for upscaling method. A proper...
We experimentally demonstrate dynamical fractal chaos in monolithic silicon optomechanical cavities, achieved through intricate intracavity coupling between radiation pressure and free-carrier Drude plasma. Chaos is measured with the correlation dimensions, Lyapunov exponents, K-entropies respectively.