Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userBCNN: Binary Complex Neural Network
72
0
0.0
(
0
)
Added by
Ang Li
Publication date
2021
fields
Informatics Engineering
and research's language is
English
Ask ChatGPT about the research
No Arabic abstract
Binarized neural networks, or BNNs, show great promise in edge-side applications with resource limited hardware, but raise the concerns of reduced accuracy. Motivated by the complex neural networks, in this paper we introduce complex representation into the BNNs and propose Binary complex neural network -- a novel network design that processes binary complex inputs and weights through complex convolution, but still can harvest the extraordinary computation efficiency of BNNs. To ensure fast convergence rate, we propose novel BCNN based batch normalization function and weight initialization function. Experimental results on Cifar10 and ImageNet using state-of-the-art network models (e.g., ResNet, ResNetE and NIN) show that BCNN can achieve better accuracy compared to the original BNN models. BCNN improves BNN by strengthening its learning capability through complex representation and extending its applicability to complex-valued input data. The source code of BCNN will be released on GitHub.
rate research
Read More
The millimeter-wave (mm-wave) radio-over-fiber (RoF) systems have been widely studied as promising solutions to deliver high-speed wireless signals to end users, and neural networks have been studied to solve various linear and nonlinear impairments. However, high computation cost and large amounts of training data are required to effectively improve the system performance. In this paper, we propose and demonstrate highly computation efficient convolutional neural network (CNN) and binary convolutional neural network (BCNN) based decision schemes to solve these limitations. The proposed CNN and BCNN based decision schemes are demonstrated in a 5 Gbps 60 GHz RoF system for up to 20 km fiber distance. Compared with previously demonstrated neural networks, results show that the bit error rate (BER) performance and the computation intensive training process are improved. The number of training iterations needed is reduced by about 50 % and the amount of required training data is reduced by over 30 %. In addition, only one training is required for the entire measured received optical power range over 3.5 dB in the proposed CNN and BCNN schemes, to further reduce the computation cost of implementing neural networks decision schemes in mm-wave RoF systems.
Being able to learn from complex data with phase information is imperative for many signal processing applications. Today s real-valued deep neural networks (DNNs) have shown efficiency in latent information analysis but fall short when applied to the complex domain. Deep complex networks (DCN), in contrast, can learn from complex data, but have high computational costs; therefore, they cannot satisfy the instant decision-making requirements of many deployable systems dealing with short observations or short signal bursts. Recent, Binarized Complex Neural Network (BCNN), which integrates DCNs with binarized neural networks (BNN), shows great potential in classifying complex data in real-time. In this paper, we propose a structural pruning based accelerator of BCNN, which is able to provide more than 5000 frames/s inference throughput on edge devices. The high performance comes from both the algorithm and hardware sides. On the algorithm side, we conduct structural pruning to the original BCNN models and obtain 20 $times$ pruning rates with negligible accuracy loss; on the hardware side, we propose a novel 2D convolution operation accelerator for the binary complex neural network. Experimental results show that the proposed design works with over 90% utilization and is able to achieve the inference throughput of 5882 frames/s and 4938 frames/s for complex NIN-Net and ResNet-18 using CIFAR-10 dataset and Alveo U280 Board.
Neural networks are discrete entities: subdivided into discrete layers and parametrized by weights which are iteratively optimized via difference equations. Recent work proposes networks with layer outputs which are no longer quantized but are solutions of an ordinary differential equation (ODE); however, these networks are still optimized via discrete methods (e.g. gradient descent). In this paper, we explore a different direction: namely, we propose a novel framework for learning in which the parameters themselves are solutions of ODEs. By viewing the optimization process as the evolution of a port-Hamiltonian system, we can ensure convergence to a minimum of the objective function. Numerical experiments have been performed to show the validity and effectiveness of the proposed methods.
This paper is concerned with the utilization of deterministically modeled chemical reaction networks for the implementation of (feed-forward) neural networks. We develop a general mathematical framework and prove that the ordinary differential equations (ODEs) associated with certain reaction network implementations of neural networks have desirable properties including (i) existence of unique positive fixed points that are smooth in the parameters of the model (necessary for gradient descent), and (ii) fast convergence to the fixed point regardless of initial condition (necessary for efficient implementation). We do so by first making a connection between neural networks and fixed points for systems of ODEs, and then by constructing reaction networks with the correct associated set of ODEs. We demonstrate the theory by constructing a reaction network that implements a neural network with a smoothed ReLU activation function, though we also demonstrate how to generalize the construction to allow for other activation functions (each with the desirable properties listed previously). As there are multiple types of networks utilized in this paper, we also give a careful introduction to both reaction networks and neural networks, in order to disambiguate the overlapping vocabulary in the two settings and to clearly highlight the role of each networks properties.
Each year a growing number of wind farms are being added to power grids to generate electricity. The power curve of a wind turbine, which exhibits the relationship between generated power and wind speed, plays a major role in assessing the performance of a wind farm. Neural networks have been used for power curve estimation. However, they do not produce a confidence measure for their output, unless computationally prohibitive Bayesian methods are used. In this paper, a probabilistic neural network with Monte Carlo dropout is considered to quantify the model (epistemic) uncertainty of the power curve estimation. This approach offers a minimal increase in computational complexity over deterministic approaches. Furthermore, by incorporating a probabilistic loss function, the noise or aleatoric uncertainty in the data is estimated. The developed network captures both model and noise uncertainty which is found to be useful tools in assessing performance. Also, the developed network is compared with existing ones across a public domain dataset showing superior performance in terms of prediction accuracy.
suggested questions
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
