- 标题
- 摘要
- 关键词
- 实验方案
- 产品
-
Unsupervised Learning Based Fast Beamforming Design for Downlink MIMO
摘要： In the downlink transmission scenario, power allocation and beamforming design at the transmitter are essential when using multiple antenna arrays. This paper considers a multiple input-multiple output broadcast channel to maximize the weighted sum-rate under the total power constraint. The classical weighted minimum mean-square error (WMMSE) algorithm can obtain suboptimal solutions but involves high computational complexity. To reduce this complexity, we propose a fast beamforming design method using unsupervised learning, which trains the deep neural network (DNN) offline and provides real-time service online only with simple neural network operations. The training process is based on an end-to-end method without labeled samples avoiding the complicated process of obtaining labels. Moreover, we use the 'APoZ'-based pruning algorithm to compress the network volume, which further reduces the computational complexity and volume of the DNN, making it more suitable for low computation-capacity devices. Finally, experimental results demonstrate that the proposed method improves computational speed significantly with performance close to the WMMSE algorithm.
关键词： beamforming,unsupervised learning,deep learning,network pruning,MIMO
更新于2025-09-23 15:23:52
-
[IEEE 2018 24th International Conference on Pattern Recognition (ICPR) - Beijing (2018.8.20-2018.8.24)] 2018 24th International Conference on Pattern Recognition (ICPR) - An Extensive Study of Cycle-Consistent Generative Networks for Image-to-Image Translation
摘要： Image-to-image translation between different domains has been an important research direction, with the aim of arbitrarily manipulating the source image content to become similar to a target image. Recently, cycle-consistent generative network (CycleGAN) has become a fundamental approach for general-purpose image-to-image translation, while almost no work has examined what factors may influence its performance. To provide more insights, we propose two new models roughly based on CycleGAN, namely LongCycleGAN and NestCycleGAN. First, LongCycleGAN cascades several generators to perform the domain translation in a long cycle. It shows the benefit of stacking more generators on the generation quality. In addition to the long cycle, NestCycleGAN develops new inner cycles to bridge intermediate generators directly, which can help constrain the unsupervised mappings. In the experiments, we conduct qualitative and quantitative comparisons for tasks including photo?label, photo?sketch, and photo colorization. The quantitative and qualitative results demonstrate the effectiveness of our two proposed models.
关键词： CycleGAN,NestCycleGAN,unsupervised learning,image-to-image translation,cycle-consistent generative networks,LongCycleGAN
更新于2025-09-23 15:22:29
-
Artificial Intelligence Algorithm Enabled Industrial-Scale Graphene Characterization
摘要： No characterization method is available to quickly perform quality inspection of 2D materials produced on an industrial scale. This hinders the adoption of 2D materials for product manufacturing in many industries. Here, we report an artificial-intelligence-assisted Raman analysis to quickly probe the quality of centimeter-large graphene samples in a non-destructive manner. Chemical vapor deposition of graphene is devised in this work such that two types of samples were obtained: layer-plus-islands and layer-by-layer graphene films, at centimeter scales. Using these samples, we implemented and integrated an unsupervised learning algorithm with an automated Raman spectroscopy to precisely cluster 20,250 and 18,000 Raman spectra collected from layer-plus-islands and layer-by-layer graphene films, respectively, into five and two clusters. Each cluster represents graphene patches with different layer numbers and stacking orders. For instance, the two clusters detected in layer-by-layer graphene films represent monolayer and bilayer graphene based on their Raman fingerprints. Our intelligent Raman analysis is fully automated, with no human operation involved, is highly reliable (99.95% accuracy), and can be generalized to other 2D materials, paving the way towards industrialization of 2D materials for various applications in the future.
关键词： unsupervised learning,graphene,two-dimensional materials,Raman spectroscopy
更新于2025-09-23 15:21:01
-
[IEEE 2019 Compound Semiconductor Week (CSW) - Nara, Japan (2019.5.19-2019.5.23)] 2019 Compound Semiconductor Week (CSW) - Radiative and Nonradiative Tunneling in Nanowire Light-Emitting Diodes
摘要： This paper aims to highlight distinctive features of the SP theory of intelligence, realized in the SP computer model, and its apparent advantages compared with some AI-related alternatives. Perhaps most importantly, the theory simplifies and integrates observations and concepts in AI-related areas, and has potential to simplify and integrate of structures and processes in computing systems. Unlike most other AI-related theories, the SP theory is itself a theory of computing, which can be the basis for new architectures for computers. Fundamental in the theory is information compression via the matching and unification of patterns and, more specifically, via a concept of multiple alignment. The theory promotes transparency in the representation and processing of knowledge, and unsupervised learning of natural structures via information compression. It provides an interpretation of aspects of mathematics and an interpretation of phenomena in human perception and cognition. Abstract concepts in the theory may be realized in terms of neurons and their inter-connections (SP-neural). These features and advantages of the SP system are discussed in relation to AI-related alternatives: the concept of minimum length encoding and related concepts, how computational and energy efficiency in computing may be achieved, deep learning in neural networks, unified theories of cognition and related research, universal search, Bayesian networks and some other models for AI, IBM’s Watson, solving problems associated with big data and in the development of intelligence in autonomous robots, pattern recognition and vision, the learning and processing of natural language, exact and inexact forms of reasoning, representation and processing of diverse forms of knowledge, and software engineering. In conclusion, the SP system can provide a firm foundation for the long-term development of AI and related areas, and at the same time, it may deliver useful results on relatively short timescales.
关键词： information compression,unsupervised learning,perception,reasoning,multiple alignment,cognition,deep learning,mathematics,neural networks,Artificial intelligence
更新于2025-09-23 15:19:57
-
[Lecture Notes in Computer Science] Computer Vision – ECCV 2018 Workshops Volume 11134 (Munich, Germany, September 8-14, 2018, Proceedings, Part VI) || Unsupervised Event-Based Optical Flow Using Motion Compensation
摘要： In this work, we propose a novel framework for unsupervised learning for event cameras that learns to predict optical flow from only the event stream. In particular, we propose an input representation of the events in the form of a discretized 3D volume, which we pass through a neural network to predict the optical flow for each event. This optical flow is used to attempt to remove any motion blur in the event image. We then propose a loss function applied to the motion compensated event image that measures the motion blur in this image. We evaluate this network on the Multi Vehicle Stereo Event Camera dataset (MVSEC), along with qualitative results from a variety of different scenes.
关键词： Optical flow,Unsupervised learning,Event cameras
更新于2025-09-19 17:15:36
-
[IEEE 2019 IEEE 46th Photovoltaic Specialists Conference (PVSC) - Chicago, IL, USA (2019.6.16-2019.6.21)] 2019 IEEE 46th Photovoltaic Specialists Conference (PVSC) - Electroluminescence-Testing Induced Crack Closure in PV Modules
摘要： This paper aims to highlight distinctive features of the SP theory of intelligence, realized in the SP computer model, and its apparent advantages compared with some AI-related alternatives. Perhaps most importantly, the theory simplifies and integrates observations and concepts in AI-related areas, and has potential to simplify and integrate of structures and processes in computing systems. Unlike most other AI-related theories, the SP theory is itself a theory of computing, which can be the basis for new architectures for computers. Fundamental in the theory is information compression via the matching and unification of patterns and, more specifically, via a concept of multiple alignment. The theory promotes transparency in the representation and processing of knowledge, and unsupervised learning of natural structures via information compression. It provides an interpretation of aspects of mathematics and an interpretation of phenomena in human perception and cognition. Abstract concepts in the theory may be realized in terms of neurons and their inter-connections (SP-neural). These features and advantages of the SP system are discussed in relation to AI-related alternatives: the concept of minimum length encoding and related concepts, how computational and energy efficiency in computing may be achieved, deep learning in neural networks, unified theories of cognition and related research, universal search, Bayesian networks and some other models for AI, IBM’s Watson, solving problems associated with big data and in the development of intelligence in autonomous robots, pattern recognition and vision, the learning and processing of natural language, exact and inexact forms of reasoning, representation and processing of diverse forms of knowledge, and software engineering. In conclusion, the SP system can provide a firm foundation for the long-term development of AI and related areas, and at the same time, it may deliver useful results on relatively short timescales.
关键词： information compression,unsupervised learning,perception,reasoning,multiple alignment,cognition,deep learning,mathematics,neural networks,Artificial intelligence
更新于2025-09-19 17:13:59
-
A Global Maximum Power Point Tracking Algorithm for Photovoltaic Systems Under Partially Shaded Conditions Using Modified Maximum Power Trapezium Method
摘要： This paper is about how the SP theory of intelligence and its realization in the SP machine (both outlined in this paper) may help in the design of the brains of autonomous robots, meaning robots that do not depend on external intelligence or power supplies, are mobile, and have human-like versatility and adaptability in intelligence. This paper addresses three main problems: 1) how to increase the computational and energy efficiency of computers and to reduce their size and weight; 2) how to achieve human-like versatility in intelligence; and 3) likewise for human-like adaptability in intelligence. Regarding the first problem, the SP system has the potential for substantial gains in computational efficiency, with corresponding cuts in energy consumption and the bulkiness of computers: 1) by reducing the size of data to be processed; 2) by exploiting statistical information that the system gathers as an integral part of how it works; and 3) via a new version of Donald Hebb’s concept of a cell assembly. Toward human-like versatility in intelligence, the SP system has strengths in unsupervised learning, natural language processing, pattern recognition, information retrieval, several kinds of reasoning, planning, problem solving, and more, with seamless integration among structures and functions. The SP system’s strengths in unsupervised learning and other aspects of intelligence may help in achieving human-like adaptability in intelligence via: 1) one-trial learning; 2) learning of natural language; 3) learning to see; 4) building 3-D models of objects and of a robot’s surroundings; 5) learning regularities in the workings of a robot and in the robot’s environment; 6) exploration and play; 7) learning major skills; and 8) learning via demonstration. Also discussed are how the SP system may process parallel streams of information, generalization of knowledge, correction of over-generalizations, learning from dirty data, how to cut the cost of learning, and reinforcements and motivations.
关键词： data compression,pattern recognition,robots,unsupervised learning,Artificial intelligence,cognitive science
更新于2025-09-19 17:13:59
-
Machine learning analysis of extreme events in optical fibre modulation instability
摘要： A central research area in nonlinear science is the study of instabilities that drive extreme events. Unfortunately, techniques for measuring such phenomena often provide only partial characterisation. For example, real-time studies of instabilities in nonlinear optics frequently use only spectral data, limiting knowledge of associated temporal properties. Here, we show how machine learning can overcome this restriction to study time-domain properties of optical fibre modulation instability based only on spectral intensity measurements. Specifically, a supervised neural network is trained to correlate the spectral and temporal properties of modulation instability using simulations, and then applied to analyse high dynamic range experimental spectra to yield the probability distribution for the highest temporal peaks in the instability field. We also use unsupervised learning to classify noisy modulation instability spectra into subsets associated with distinct temporal dynamic structures. These results open novel perspectives in all systems exhibiting instability where direct time-domain observations are difficult.
关键词： machine learning,extreme events,optical fibre modulation instability,unsupervised learning,supervised neural network
更新于2025-09-10 09:29:36
-
[Lecture Notes in Computer Science] Advances in Visual Computing Volume 11241 (13th International Symposium, ISVC 2018, Las Vegas, NV, USA, November 19 – 21, 2018, Proceedings) || Road User Abnormal Trajectory Detection Using a Deep Autoencoder
摘要： In this paper, we focus on the development of a method that detects abnormal trajectories of road users at tra?c intersections. The main di?culty with this is the fact that there are very few abnormal data and the normal ones are insu?cient for the training of any kinds of machine learning model. To tackle these problems, we proposed the solution of using a deep autoencoder network trained solely through augmented data considered as normal. By generating arti?cial abnormal trajectories, our method is tested on four di?erent outdoor urban users scenes and performs better compared to some classical outlier detection methods.
关键词： Deep autoencoder,Data augmentation,Abnormal trajectory detection,Unsupervised learning
更新于2025-09-10 09:29:36
-
[IEEE 2018 International Joint Conference on Neural Networks (IJCNN) - Rio de Janeiro (2018.7.8-2018.7.13)] 2018 International Joint Conference on Neural Networks (IJCNN) - STDP Learning of Image Patches with Convolutional Spiking Neural Networks
摘要： Spiking neural networks are motivated from principles of neural systems and may possess unexplored advantages in the context of machine learning. A class of convolutional spiking neural networks is introduced, trained to detect image features with an unsupervised, competitive learning mechanism. Image features can be shared within subpopulations of neurons, or each may evolve independently to capture different features in different regions of input space. We analyze the time and memory requirements of learning with and operating such networks. The MNIST dataset is used as an experimental testbed, and comparisons are made between the performance and convergence speed of a baseline spiking neural network.
关键词： Spiking Neural Networks,Unsupervised Learning,Convolution,STDP,Machine Learning
更新于2025-09-09 09:28:46