scholarly journals A Systolic Accelerator for Neuromorphic Visual Recognition

Electronics ◽  
2020 ◽  
Vol 9 (10) ◽  
pp. 1690
Author(s):  
Shuo Tian ◽  
Lei Wang ◽  
Shi Xu ◽  
Shasha Guo ◽  
Zhijie Yang ◽  
...  
Keyword(s):  
Visual Recognition ◽  
Computational Models ◽  
Partial Sums ◽  
Memory Bandwidth ◽  
Two Dimensional ◽  
Biologically Inspired ◽  
Speed Up ◽  
Simulation Results ◽  
S2 Layer ◽  
Weight Data

Advances in neuroscience have encouraged researchers to focus on developing computational models that behave like the human brain. HMAX is one of the potential biologically inspired models that mimic the primate visual cortex’s functions and structures. HMAX has shown its effectiveness and versatility in multi-class object recognition with a simple computational structure. It is still a challenge to implement the HMAX model in embedded systems due to the heaviest computational S2 phase of HMAX. Previous implementations such as CoRe16 have used a reconfigurable two-dimensional processing element (PE) array to speed up the S2 layer for HMAX. However, the adder tree mechanism in CoRe16 used to produce output pixels by accumulating partial sums in different PEs increases the runtime for HMAX. To speed up the execution process of the S2 layer in HMAX, in this paper, we propose SAFA (systolic accelerator for HMAX), a systolic-array based architecture to compute and accelerate the S2 stage of HMAX. Using the output stationary (OS) dataflow, each PE in SAFA not only calculates the output pixel independently without additional accumulation of partial sums in multiple PEs, but also reduces the multiplexers applied in reconfigurable accelerators. Besides, data forwarding for the same input or weight data in OS reduces the memory bandwidth requirements. The simulation results show that the runtime of the heaviest computational S2 stage in HMAX model is decreased by 5.7%, and the bandwidth required for memory is reduced by 3.53 × on average by different kernel sizes (except for kernel = 12) compared with CoRe16. SAFA also obtains lower power and area costs than other reconfigurable accelerators from synthesis on ASIC.

Design of finite-time guidance law based on observer and head-pursuit theory

2021 ◽  
pp. 095441002098456
Author(s):  
Chenqi Zhu
Keyword(s):  
Finite Time ◽  
Three Dimensional ◽  
Hypersonic Vehicle ◽  
Two Dimensional ◽  
Dimensional Model ◽  
Reaching Law ◽  
Guidance Law ◽  
Fast Power ◽  
Simulation Results ◽  
Guidance Laws

In order to improve the guiding accuracy in intercepting the hypersonic vehicle, this article presents a finite-time guidance law based on the observer and head-pursuit theory. First, based on a two-dimensional model between the interceptor and target, this study applies the fast power reaching law to head-pursuit guidance law so that it can alleviate the chattering phenomenon and ensure the convergence speed. Second, target maneuvers are considered as system disturbances, and the head-pursuit guidance law based on an observer is proposed. Furthermore, this method is extended to a three-dimensional case. Finally, comparative simulation results further verify the superiority of the guidance laws designed in this article.

Computational models of filtration gas combustion

2017 ◽  
Vol 32 (2) ◽  
Author(s):  
Yuri M. Laevsky ◽  
Tatyana A. Nosova
Keyword(s):  
Heat Flow ◽  
Numerical Experiments ◽  
Computational Models ◽  
Two Dimensional ◽  
Multidimensional Model ◽  
Gas Combustion ◽  
Total Enthalpy ◽  
Diffusive Flow ◽  
Temperature Heat ◽  
The One

AbstractA multidimensional model of filtration gas combustion is presented. The model is based on the system of conservation laws of ‘temperature – heat flow’, ‘mass–diffusive flow’ types with introducing the concept of total enthalpy flow. Results of numerical experiments are presented for the one- and two-dimensional problems for different conditions and parameters.

Direct simulation of evolution and control of three-dimensional instabilities in attachment-line boundary layers

1995 ◽  
Vol 291 ◽  
pp. 369-392 ◽  
Author(s):  
Ronald D. Joslin
Keyword(s):  
Boundary Layer ◽  
Stokes Equations ◽  
Three Dimensional ◽  
Plate Boundary ◽  
Navier Stokes ◽  
Two Dimensional ◽  
Computationally Efficient ◽  
Simulation Results ◽  
Dimensional Simulation ◽  
Attachment Line

The spatial evolution of three-dimensional disturbances in an attachment-line boundary layer is computed by direct numerical simulation of the unsteady, incompressible Navier–Stokes equations. Disturbances are introduced into the boundary layer by harmonic sources that involve unsteady suction and blowing through the wall. Various harmonic-source generators are implemented on or near the attachment line, and the disturbance evolutions are compared. Previous two-dimensional simulation results and nonparallel theory are compared with the present results. The three-dimensional simulation results for disturbances with quasi-two-dimensional features indicate growth rates of only a few percent larger than pure two-dimensional results; however, the results are close enough to enable the use of the more computationally efficient, two-dimensional approach. However, true three-dimensional disturbances are more likely in practice and are more stable than two-dimensional disturbances. Disturbances generated off (but near) the attachment line spread both away from and toward the attachment line as they evolve. The evolution pattern is comparable to wave packets in flat-plate boundary-layer flows. Suction stabilizes the quasi-two-dimensional attachment-line instabilities, and blowing destabilizes these instabilities; these results qualitatively agree with the theory. Furthermore, suction stabilizes the disturbances that develop off the attachment line. Clearly, disturbances that are generated near the attachment line can supply energy to attachment-line instabilities, but suction can be used to stabilize these instabilities.

Influence of Engine Cabin Simplification on Aerodynamic Characteristics of Car

2014 ◽  
Vol 1042 ◽  
pp. 188-193 ◽  
Author(s):  
Xing Jun Hu ◽  
Jing Chang
Keyword(s):  
Numerical Simulation ◽  
Aerodynamic Drag ◽  
Aerodynamic Characteristics ◽  
Thin Plates ◽  
Average Thickness ◽  
Two Dimensional ◽  
Engine Cooling ◽  
Aerodynamic Drag Coefficient ◽  
Simulation Results ◽  
The Impact

In order to analyze the impact of engine cabin parts on aerodynamic characteristics, the related parts are divided into three categories except the engine cooling components: front thin plates (average thickness of 2mm), bottom-suspension and interior panels. The aerodynamic drag coefficient (Cd) were obtained upon the combination schemes consisting of the three types of parts by numerical simulation. Results show that Cd by simulation is closer to the test value gained by the wind tunnel experiment when front thin plates were simplified to the two-dimensional interface with zero thickness. The error is only 5.23%. Meanwhile this scheme reduces grid numbers, thus decreasing the calculating time. As the front thin plates can guide the flow, there is no difference on the Cd values gained from the model with or without bottom-suspension or interior panels when the engine cabin contains the front thin plates; while only both bottom-suspension and interior panels are removed, the Cd value can be reduced when the cabin doesn’t contain the front thin plates.

scholarly journals Numerical modeling of microwave heating

2010 ◽  
Vol 42 (1) ◽  
pp. 99-124 ◽  
Author(s):  
A.K. Shukla ◽  
A. Mondal ◽  
A. Upadhyaya
Keyword(s):  
Thermal Conductivity ◽  
Numerical Modeling ◽  
Temperature Distribution ◽  
Microwave Heating ◽  
Two Dimensional ◽  
Conventional Furnace ◽  
Microwave Furnace ◽  
Simulation Results ◽  
Heating Behavior ◽  
Cylindrical Samples

The present study compares the temperature distribution within cylindrical samples heated in microwave furnace with those achieved in radiatively-heated (conventional) furnace. Using a two-dimensional finite difference approach the thermal profiles were simulated for cylinders of varying radii (0.65, 6.5, and 65 cm) and physical properties. The influence of susceptor-assisted microwave heating was also modeled for the same. The simulation results reveal differences in the heating behavior of samples in microwaves. The efficacy of microwave heating depends on the sample size and its thermal conductivity.

scholarly journals A neuronal circuit for vector computation builds an allocentric traveling-direction signal in the Drosophila fan-shaped body

2020 ◽  
Author(s):  
Cheng Lyu ◽  
L.F. Abbott ◽  
Gaby Maimon
Keyword(s):  
Computational Models ◽  
Neuronal Population ◽  
The Body ◽  
Central Complex ◽  
Two Dimensional ◽  
Neuronal Circuit ◽  
Head Direction ◽  
Neural Signals ◽  
Shaped Body ◽  
External Cues

AbstractMany behavioral tasks require the manipulation of mathematical vectors, but, outside of computational models1–8, it is not known how brains perform vector operations. Here we show how the Drosophila central complex, a region implicated in goal-directed navigation8–14, performs vector arithmetic. First, we describe neural signals in the fan-shaped body that explicitly track a fly’s allocentric traveling direction, that is, the traveling direction in reference to external cues. Past work has identified neurons in Drosophila12,15–17 and mammals18,19 that track allocentric heading (e.g., head-direction cells), but these new signals illuminate how the sense of space is properly updated when traveling and heading angles differ. We then characterize a neuronal circuit that rotates, scales, and adds four vectors related to the fly’s egocentric traveling direction–– the traveling angle referenced to the body axis––to compute the allocentric traveling direction. Each two-dimensional vector is explicitly represented by a sinusoidal activity pattern across a distinct neuronal population, with the sinusoid’s amplitude representing the vector’s length and its phase representing the vector’s angle. The principles of this circuit, which performs an egocentric-to-allocentric coordinate transformation, may generalize to other brains and to domains beyond navigation where vector operations or reference-frame transformations are required.

Dual-mode of topological rainbow in gradual photonic heterostructures

2021 ◽  
Author(s):  
Mingxing Li ◽  
Yueke Wang ◽  
Mengjia Lu ◽  
Tian Sang
Keyword(s):  
Finite Element Method ◽  
Finite Element Method Simulation ◽  
Two Dimensional ◽  
Edge States ◽  
Dual Mode ◽  
Strong Robustness ◽  
Mode Excitation ◽  
Simulation Results ◽  
Photonic States ◽  
Sandwiched Structure

Abstract In this letter, a method to realize the topological rainbow trapping is presented, which is composed of gradual ordinary-topological-ordinary heterostructures based on two-dimensional photonic crystals with C-4 symmetry. In the proposed sandwiched structure, the two coupled topological edge states with different frequencies are separated and trapped in different positions, due to group velocity of near to zero. We have achieved the dual-mode of topological rainbow in one structure, which broadens the bandwidth. Besides, the dual-mode of topological rainbow under one mode excitation is also realized by using a simple bend design. The immunity to defects is also investigated and it is found our slowing light system has strong robustness. Finite Element Method simulation results verify our idea, and our work opens up a new way for frequency routing and broadband operation of topological photonic states.

Machine Learning Applications in Computer Vision

2013 ◽  
pp. 896-926
Author(s):  
Mehrtash Harandi ◽  
Javid Taheri ◽  
Brian C. Lovell
Keyword(s):  
Machine Learning ◽  
Computer Vision ◽  
Object Recognition ◽  
Visual Recognition ◽  
Sequential Decision ◽  
Biologically Inspired ◽  
Transductive Learning ◽  
Machine Learning Applications ◽  
Appearance Models ◽  
Learning Frameworks

Recognizing objects based on their appearance (visual recognition) is one of the most significant abilities of many living creatures. In this study, recent advances in the area of automated object recognition are reviewed; the authors specifically look into several learning frameworks to discuss how they can be utilized in solving object recognition paradigms. This includes reinforcement learning, a biologically-inspired machine learning technique to solve sequential decision problems and transductive learning, and a framework where the learner observes query data and potentially exploits its structure for classification. The authors also discuss local and global appearance models for object recognition, as well as how similarities between objects can be learnt and evaluated.

An All Optical OR Gate Using BPSK Technique Inside 2D Photonic Crystals

2019 ◽  
Vol 0 (0) ◽  
Author(s):  
Bahman Shahmohammadi ◽  
Alireza Andalib
Keyword(s):  
Phase Shift ◽  
Photonic Crystals ◽  
Destructive Interference ◽  
Two Dimensional ◽  
Phase Shift Keying ◽  
All Optical ◽  
Shift Keying ◽  
Working Principle ◽  
Simulation Results ◽  
Or Gate

AbstractIn this paper, we are going to design simulate an all optical OR gate based on two-dimensional photonic crystals. The proposed structure was designed based on phase shift keying technique. The working principle of the proposed structure is based on constructive and destructive interference of optical beams. The simulation results shows that the rise and fall times are 0.8 and 0.2 ps, respectively.

Sign in / Sign up

Export Citation Format

Share Document