5 results about "Chaotic neural network" patented technology

The application relates to the technical field of signal processing, in particular to a chaotic neural network with complex-valued weights and application of the chaotic neural network in electrocardiogram classification. First, a complex Logistic chaotic mapping with complex-valued parameters is proposed, and bifurcation diagrams, Lyapunov exponents and chaotic attractors of the complex Logistic chaotic mapping are analyzed. Second, the ergodicity of CLCM and a new neuron function are used to optimize the weights of the CNN. Then, the method is verified by using the MIT-BIH database. Through band-pass filtering and double-threshold processing, the electrocardiogram signal is processed into a signal with a single waveform and a more prominent signal, which is used as the input of the designed CNN. The results show that the accuracy of the CNN with complex weights for electrocardiogram classification is improved. The complex weight chaotic neural network has the ability to prevent the network from falling into a local minimum, improves the recognition accuracy of the electrocardiogram, and has a high recognition accuracy.

This invention relates to the field of image encryption technology and addresses the problems of slow encryption speed and weak resistance to attacks in existing image encryption methods. The image encryption method described in this invention uses a ternary fractional discrete chaotic neural network system to generate a chaotic sequence related to the plaintext. A fractal sorting and scrambling process is then performed on the original image using a fractal model approach. To achieve better results, a double scrambling process is applied to both rows and columns. Finally, a synchronous scrambling and diffusion operation is performed according to the Hilbert curve traversal order, which can simultaneously change the position and size of pixel values. This method is the first to apply fractal thinking to the scrambling process in encryption, achieving good scrambling results. The synchronous scrambling and diffusion operation further improves the encryption efficiency to a certain extent.

The invention discloses a medical image content extraction method based on a trainable multilayer RBF-Chaotic neural network. According to the method, medical image content is extracted based on the trainable multilayer RBF-Chaotic neural network; the trainable multi-layer RBF-Chaotic neural network is obtained by simplifying a LeNet-5 network structure, and comprises the steps of removing all maximum pooling layers in the LeNet-5 network structure, reconstructing a classifier into a three-layer full-connection layer, and introducing chaotic momentum generated by a 2DES function in a training process to replace traditional SGD momentum. According to the method, the model is thinner, the occupation of a video memory is lower, a consumer-level video card can be completely loaded, offline clustering is not needed, the model only changes training and does not change reasoning, hospital night long-time training can be successful at a time, and the manual parameter adjustment cost is remarkably reduced.

The application belongs to the field of memristor neural network and secret communication, and particularly relates to a global projection synchronization method of a four-element value memristor neural network based on open-loop control and application thereof. The method is different from the traditional separation of the system into a real number system and three complex number systems, but the system is regarded as a whole for research. Meanwhile, a discrete multi-time delay memristor neural network is used to establish a corresponding drive system and response system, and an open-loop controller is designed, so that the transmission of the encrypted plaintext signal in the channel can achieve better security effect. Compared with the traditional chaotic neural network, the application has better security effect, lower network energy consumption, and provides a better solution for encrypted communication.

The invention discloses an astrocyte memristive network driven time sequence analysis and dynamic security method, system and device and a medium, and belongs to the technical field of computer vision and information security, and the method comprises the steps: collecting a video stream, and extracting a human skeleton sequence; performing space-time alignment evaluation on the skeleton sequence and a standard template, and identifying an abnormal action; constructing a quadruple synaptic network comprising Chay neurons, HR neurons, astrocytes and memristors, and dynamically adjusting electrical coupling parameters through gesture recognition to generate chaotic signals; extracting multi-dimensional features from the chaotic signal, and generating a key stream through Hash enhancement and key expansion; encrypting the video data by using the key stream; and generating a scoring suggestion. Lightweight real-time encryption is realized through the biologically inspired chaotic neural network, the problem of high calculation complexity of a traditional encryption algorithm is solved, real-time performance and safety are both considered, gesture dynamic control is supported, and the method is suitable for scenes such as rehabilitation training and posture correction.