Figure 4 - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
Source publication
Optical neural networks can effectively address hardware constraints and parallel computing efficiency issues inherent in electronic neural networks. However, the inability to implement convolutional neural networks at the all-optical level remains a hurdle. In this work, we propose an optical diffractive convolutional neural network (ODCNN) that i...
Context in source publication
Context 1
... it is not necessary to consider the output form of the feature map, which can directly process through the fully connected layer, whether stacked or tiled. Figure 4 shows the structure of the fully connected layer. The fully connected layer is composed of a diffractive deep neural network. ...
Similar publications
Due to the inherent insecure nature of the Internet, it is crucial to ensure the secure transmission of image data over this network. Additionally, given the limitations of computers, it becomes even more important to employ efficient and fast image encryption techniques. While 1D chaotic maps offer a practical approach to real-time image encryptio...
Citations
... However, the integration of optical and electronic devices is challenging due to their different size scales [16]. Consequently, all-optical neural networks have been developed [17][18][19][20][21], offering higher integration and lower energy consumption. This has significant implications for applications like spacecraft, satellites, and space stations, where computational power is severely limited [22,23]. ...
Compared to traditional neural networks, optical neural networks demonstrate significant advantages in terms of information processing speed, energy efficiency, anti-interference capability, and scalability. Despite the rapid development of optical neural networks in recent years, most existing systems still face challenges such as complex structures, time-consuming training, and insufficient accuracy. This study fully leverages the coherence of optical systems and introduces an optical Fourier convolutional neural network based on the diffraction of complex image light fields. This new network is not only structurally simple and fast in computation but also excels in image classification accuracy. Our research opens new perspectives for the development of optical neural networks, and also offers insights for future applications in high-efficiency, low-energy-consumption computing domains.
... It was also proposed to use phase transitions, which are expressed in in the form of sharp solution turbidity, for the creation of information display systems [16][17][18][19][20][21]. Currently, this issue is being actualized in connection with attempts to achieve optical neural networks with controlled characteristics of optical elements [22][23][24][25]. considered copolymers and PAA is influenced not only by the formation of hydrophilic interpolymer associates (HIAs) studied in [45], but also by the formation of HHAs, the existence of which was proven in [46]. ...
Citation: Shaikhutdinov, R.; Mun, G.; Kopishev, E.; Bakirov, A.; Kabdushev, S.; Baipakbaeva, S.; Suleimenov, I. Effect of the Formation of Hydrophilic and Hydrophobic-Hydrophilic Associates on the Behavior of Copolymers of N-Vinylpyrrolidone with Methyl Acrylate in Aqueous Solutions. Polymers 2024, 16, 584. Abstract: It has been shown that there exist conditions under which thermosensitive copolymers of N-vinylpyrrolidone with methyl acrylate form hydrophobic-hydrophilic associations, which are unstable dynamic meshes, the bonds in which are continuously broken and created again, and the nature of the formation of such meshes depends significantly on the proportion of the hydrophobic component in the copolymer. It is shown that the interaction of the above copolymers with polyacrylic acid results in the formation of not only classical interpolymer complexes, but also hydrophilic interpolymer associates, which also represent unstable networks existing in a dynamic mode. In such meshes, the molecules of the above copolymers serve as a kind of cross-agent connecting the polyacid molecules. There are also conditions under which such meshes acquire a complex structure, since unstable bonds between macromolecular tangles of both the same and different types take part in their formation. It is shown that the transition from the formation of interpolymer complexes to the formation of hydrophilic interpolymer associates can occur, among other things, due to changes in the acidity or concentration of low-molecular salt in solution.
... This system is called a multiwavelength D 2 NN [61]. Yu et al. proposed an optical diffraction convolutional neural network (ODCNN) based on an optical 4 f system and diffractive fully connected layers [62]. Li et al. demonstrated improvements to the diffractive optical neural network using differential measurement techniques, which alleviated the strict non-negativity constraints on the light intensity [63]. ...
In 2018, a UCLA research group published an important paper on optical neural network (ONN) research in the journal Science. It developed the world’s first all-optical diffraction deep neural network (DNN) system, which can perform MNIST dataset classification tasks at near-light-speed. To be specific, the UCLA research group adopted a terahertz light source as the input, established the all-optical diffractive DNN ( ${{\rm{D}}^2}{\rm{NN}}$ D 2 N N ) model using the Rayleigh-Sommerfeld diffraction theory, optimized the model parameters using the stochastic gradient descent algorithm, and then used 3D printing technology to make the diffraction grating and built the ${{\rm{D}}^2}{\rm{NN}}$ D 2 N N system. This research opened a new ONN research direction. Here, we first review and analyze the development history and basic theory of artificial neural networks (ANNs) and ONNs. Second, we elaborate ${{\rm{D}}^2}{\rm{NN}}$ D 2 N N as holographic optical elements (HOEs) interconnected by free space light and describe the theory of ${{\rm{D}}^2}{\rm{NN}}$ D 2 N N . Then we cover the nonlinear research and application scenarios for ${{\rm{D}}^2}{\rm{NN}}$ D 2 N N . Finally, the future directions and challenges of ${{\rm{D}}^2}{\rm{NN}}$ D 2 N N are briefly discussed. Hopefully, our work can provide support and help to researchers who study the theory and application of ${{\rm{D}}^2}{\rm{NN}}$ D 2 N N in the future.
