A recurrent stochastic binary network

Stochastic neural networks are usually built by introducing random fluctuations into the network. A natural method is to use stochastic connections rather than stochastic activation functions. We propose a new model in which each neuron has very simple functionality but all the connections are stochastic. It is shown that the stationary distribution of the network uniquely exists and it is approxi-mately a Boltzmann-Gibbs distribution. The relationship between the model and the Markov random field is discussed. New techniques to implement simulated annealing and Boltzmann learning are pro-posed. Simulation results on the graph bisection problem and image recognition show that the network is powerful enough to solve real world problems.

A Novel Pitch Determination Algorithm with Binary Lateral Inhibition Network

A novel pitch determination algorithm with binary lateral inhibition network (BLIN) is proposed in this paper. A thread like spectrum, which is composed of the fundamental frequency and its harmonic components, is acquired by applying a BLIN to the short time spectrum of the speech signal. Then the pitch is determined by the average interval of harmonics. The algorithm is evaluated on COSDIC speech database. For comparison, the same results obtained from the same speech sample with the cepstrum and autocorrelation based pitch determination algorithms are also presented. The results show that the new algorithm is superior to the cepstrum and autocorrelation based pitch determination algorithms.

GAAF:Searching Activation Functions for Binary Neural Networks Through Genetic Algorithm

Binary neural networks(BNNs)show promising utilization in cost and power-restricted domains such as edge devices and mobile systems.This is due to its significantly less computation and storage demand,but at the cost of degraded performance.To close the accuracy gap,in this paper we propose to add a complementary activation function(AF)ahead of the sign based binarization,and rely on the genetic algorithm(GA)to automatically search for the ideal AFs.These AFs can help extract extra information from the input data in the forward pass,while allowing improved gradient approximation in the backward pass.Fifteen novel AFs are identified through our GA-based search,while most of them show improved performance(up to 2.54%on ImageNet)when testing on different datasets and network models.Interestingly,periodic functions are identified as a key component for most of the discovered AFs,which rarely exist in human designed AFs.Our method offers a novel approach for designing general and application-specific BNN architecture.GAAF will be released on GitHub.

Area Efficient Pattern Representation of Binary Neural Networks on RRAM

Resistive random access memory(RRAM)has been demonstrated to implement multiply-and-accumulate(MAC)operations using a highly parallel analog fashion,which dramatically accelerates the convolutional neural networks(CNNs).Since CNNs require considerable converters between analog crossbars and digital peripheral circuits,recent studies map the binary neural networks(BNNs)onto RRAM and binarize the weights to{+1,-1}.However,two mainstream representations for BNN weights introduce patterns of redundant 0s and 1s when dealing with negative weights.In this work,we reduce the area of redundant 0s and 1s by proposing a BNN weight representation framework based on the novel pattern representation and a corresponding architecture.First,we spilt the weight matrix into several small matrices by clustering adjacent columns together.Second,we extract 1s'patterns,i.e.,the submatrices only containing 1s,from the small weight matrix,such that each final output can be represented by the sum of several patterns.Third,we map these patterns onto RRAM crossbars,including pattern computation crossbars(PCCs)and pattern accumulation crossbars(PACs).Finally,we compare the pattern representation with two mainstream representations and adopt the more area efficient one.The evaluation results demonstrate that our framework can save over 20%of crossbar area effectively,compared with two mainstream representations.

Reliability Analysis of Aircraft Condition Monitoring Network Using an Enhanced BDD Algorithm

The aircraft condition monitoring network is responsible for collecting the status of each component in aircraft. The reliability of this network has a significant effect on safety of the aircraft. The aircraft condition monitoring network works in a real-time manner that all the data should be transmitted within the deadline to ensure that the control center makes proper decision in time. Only the connectedness between the source node and destination cannot guarantee the data to be transmitted in time. In this paper, we take the time deadline into account and build the task-based reliability model. The binary decision diagram （BDD）, which has the merit of efficiency in computing and storage space, is introduced when calculating the reliability of the network and addressing the essential variable. A case is analyzed using the algorithm proposed in this paper. The experimental results show that our method is efficient and proper for the reliability analysis of the real-time network.

Dip-coating processed sponge-based electrodes for stretchable Zn-MnO2 batteries

Stretchable electronics are in high demand for next-generation wearable devices, but their fabrication is still challenging. Stretchable conductors, flexible pressure sensors, and foldable light-emitting diodes （LEDs） have been reported; however, the fabrication of stable stretchable batteries, as power suppliers for wearable devices, is significantly behind the development of other stretchable electronics. Several stretchable lithium-ion batteries and primary batteries have been fabricated, but their low capacities and complicated manufacturing processes are obstacles for practical applications. Herein, we report a stretchable zinc/manganese-oxide （Zn-MnO2） full battery based on a silver-nanowire- coated sponge prepared via a facile dip-coating process. The spongy electrode, with a three-dimensional （3D） binary network structure, provided not only high conductivity and stretchability, but also enabled a high mass loading of electrochemically active materials （Zn and MnO2 particles）. The fabricated Zn-MnO2 battery exhibited an areal capacity as high as 3.6 mAh·cm^-2 and could accommodate tensile strains of up to 100% while retaining 89% of its original capacity. The facile solution-based strategy of dip-coating active materials onto a cheap sponge-based stretchable current collector opens up a new avenue for fabricating stretchable batteries.

Segregation characteristics of irregular binaries in gas solid fluidized beds——An ANN-approach

Binary mixtures of irregular materials of different particle sizes and/or particle densities are fluidized in a 15-cm diameter column with a perforated plate distributor. An attempt has been made in this work to determine the segregation characteristics of jetsam particles for both the homogeneous and heterogeneous binary mixtures in terms of segregation distance by correlating it to the various system parameters, viz. initial static bed height, height of a layer of particles above the bottom grid, superficial gas velocity and average particle size and/or particle densities of the mixture through the dimensional analysis. Correlation on the basis of Artificial Neural Network approach has also been developed with the above system parameters thereby authenticating the development of correlation by the former approach. The calculated values of the segregation distance obtained for both the homogeneous and heterogeneous binary mixtures from both the types of ftuidized beds （i.e. under the static bed condition and the ftuidized bed condition） have also been compared with each other.