In recent years, deep learning has garnered tremendous success in a variety of application domains. This new field of machine learning has been growing rapidly and has been applied to most traditional application domains, as well as some new areas that present more opportunities. Different methods have been proposed based on different categories of learning, including supervised, semi-supervised, and un-supervised learning. Experimental results show state-of-the-art performance using deep learning when compared to traditional machine learning approaches in the fields of image processing, computer vision, speech recognition, machine translation, art, medical imaging, medical information processing, robotics and control, bioinformatics, natural language processing, cybersecurity, and many others. This survey presents a brief survey on the advances that have occurred in the area of Deep Learning (DL), starting with the Deep Neural Network (DNN). The survey goes on to cover Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), including Long Short-Term Memory (LSTM) and Gated Recurrent Units (GRU), Auto-Encoder (AE), Deep Belief Network (DBN), Generative Adversarial Network (GAN), and Deep Reinforcement Learning (DRL). Additionally, we have discussed recent developments, such as advanced variant DL techniques based on these DL approaches. This work considers most of the papers published after 2012 from when the history of deep learning began. Furthermore, DL approaches that have been explored and evaluated in different application domains are also included in this survey. We also included recently developed frameworks, SDKs, and benchmark datasets that are used for implementing and evaluating deep learning approaches. There are some surveys that have been published on DL using neural networks and a survey on Reinforcement Learning (RL). However, those papers have not discussed individual advanced techniques for training large-scale deep learning models and the recently developed method of generative models.
A novel amplitude-modulated phase-only filter (AMPOF) is proposed for achieving improved correlation discrimination. The proposed AMPOF has an amplitude spectrum which is the inverse of a biased amplitude spectrum of the object function and a phase spectrum which is a complex conjugate of the phase spectrum of the object function. When compared with the phase-only filters, the AMPOF is found to have significantly superior correlation discrimination capability.
A general design algorithm is presented for the multioutput polarization-encoded optical shadow-casting scheme. A set of POSC equations is obtained from the truth table of the desired logic unit and is solved in terms of four possible pixel characteristics (transparent, opaque, vertically polarized, and horizontally polarized) and four possible source characteristics (off, unpolarized, vertically polarized, and horizontally polarized). To demonstrate its feasibility, the algorithm is used to determine the input pixel characteristics of a full adder and a full subtracter.
A system of customized spatial light modulators has been installed onto the front end of the laser system at the National Ignition Facility (NIF). The devices are capable of shaping the beam profile at a low-fluence relay plane upstream of the amplifier chain. Their primary function is to introduce "blocker" obscurations at programmed locations within the beam profile. These obscurations are positioned to shadow small, isolated flaws on downstream optical components that might otherwise limit the system operating energy. The modulators were designed to enable a drop-in retrofit of each of the 48 existing Pre Amplifier Modules (PAMs) without compromising their original performance specifications. This was accomplished by use of transmissive Optically Addressable Light Valves (OALV) based on a Bismuth Silicon Oxide photoconductive layer in series with a twisted nematic liquid crystal (LC) layer. These Programmable Spatial Shaper packages in combination with a flaw inspection system and optic registration strategy have provided a robust approach for extending the operational lifetime of high fluence laser optics on NIF.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.