Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Signal processing algorithms”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 253 records · Page 14
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Monitoring of corrosion effects in pipes with multi-mode acoustic signals

Current commercial long-range guided wave inspection and monitoring technologies for pipes usually require establishment of certain specific acoustic modes for wave propagation and use a large number of sensors in a certain geometric arrangement (for example, uniformly-spaced arrays of sensors on a collar, placed around a pipe). This paper, in contrast, presents a long-range corrosion detection method in pipes based on multi-mode guided acoustic waves, generated and received using simple, single-element sensors. In this technique, available guided wave modes in any interrogated zone of the structure are automatically down-selected by the medium, and changes in energy distribution over the selected modes are analyzed through signal processing to monitor material loss or addition. Here, to investigate the feasibility of the proposed method, we have performed both experimental and numerical simulation studies on 50.8-mm (2-in)-diameter schedule-40 carbon steel pipes. In these studies, a linear chirp signal is injected in compression mode, perpendicular to the surface of the pipe, generating multi-mode, multi-directional guided waves that travel within the pipe wall. Corrosion effects in the pipes are simulated numerically and experimentally. To reduce errors in the measurement technique due to changes in the local or global temperature of the measured media, a universal temperature compensation algorithm is introduced. The experimental results show good agreement with simulation results, indicating that the proposed multi-mode wave generation and detection in a pipe could be used for long-range detection and monitoring of corrosion effects, such as pipe wall losses or solid material deposits.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Internship Work Report

I worked on two projects during my summer internship at Sandia. My official title was “Intern - Mission Tech Electrical Eng./Computer Eng.- R&D Undergraduate Summer.” I worked at the central location, which is Albuquerque, New Mexico. The department you are placed in at Sandia doesn’t always correspond to the people you will be working with. For example, I only directly worked with one person from my department this summer. On one of my projects, I worked with a diverse team of engineers from many different departments. On my other project, I mainly worked with two departments, as the project had two distinct parts. As mentioned earlier, I worked on two projects during my summer at Sandia. The first project focused on a lightweight embedded controller in an advanced FPGA System-on-Chip for radar signal processing applications. The term “controller” refers to a hardware device that directs the flow of data between two entities. An FPGA is a reprogrammable integrated circuit (as opposed to an integrated circuit with one purpose). An FPGA was used on this project so in order to protype various ideas for our System-on-Chip. My role on the project was to implement designs on the fabric of the FPGA and design a state machine (written in C) for the processor. My second project was also heavily involved with embedded systems but had a different application. It focused on using a Newton-Raphson control algorithm to stabilize an inverted pendulum using a novel microcontroller. The pendulum dynamics were derived, and it was successfully simulated in MATLAB. I worked on integrating the microcontroller with the inverted pendulum machinery, and converting the Newton-Raphson control algorithm from MATLAB into C. The inverted pendulum was successfully stabilized using a simple PID controller and industry-standard microcontroller. The project is still ongoing, and the team is gearing up for more tests using the novel Newton-Raphson control algorithm and novel microcontroller

42 ENGINEERING↗
DOE OSTI2021

Image-Based Methods to Score Fungal Pathogen Symptom Progression and Severity in Excised Arabidopsis Leaves

Image-based symptom scoring of plant diseases is a powerful tool for associating disease resistance with plant genotypes. Advancements in technology have enabled new imaging and image processing strategies for statistical analysis of time-course experiments. There are several tools available for analyzing symptoms on leaves and fruits of crop plants, but only a few are available for the model plant Arabidopsis thaliana (Arabidopsis). Arabidopsis and the model fungus Botrytis cinerea (Botrytis) comprise a potent model pathosystem for the identification of signaling pathways conferring immunity against this broad host-range necrotrophic fungus. Here, we present two strategies to assess severity and symptom progression of Botrytis infection over time in Arabidopsis leaves. Thus, a pixel classification strategy using color hue values from red-green-blue (RGB) images and a random forest algorithm was used to establish necrotic, chlorotic, and healthy leaf areas. Secondly, using chlorophyll fluorescence (ChlFl) imaging, the maximum quantum yield of photosystem II (Fv/Fm) was determined to define diseased areas and their proportion per total leaf area. Both RGB and ChlFl imaging strategies were employed to track disease progression over time. This has provided a robust and sensitive method for detecting sensitive or resistant genetic backgrounds. A full methodological workflow, from plant culture to data analysis, is described.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

Imaging and Segmenting Grains and Subgrains Using Backscattered Electron Techniques

We present two new methods of processing data from backscattered electron signals in a scanning electron microscope to image grains and subgrains. The first combines data from multiple backscattered electron images acquired at different specimen geometries to (1) better reveal grain boundaries in recrystallized microstructures and (2) distinguish between recrystallized and unrecrystallized regions in partially recrystallized microstructures. The second utilizes spherical harmonic transform indexing of electron backscatter diffraction patterns to produce high angular resolution orientation data that enable the characterization of subgrains. Subgrains are produced during high-temperature plastic deformation and have boundary misorientation angles ranging from a few degrees down to a few hundredths of a degree. Here, we also present an algorithm to automatically segment grains from combined backscattered electron image data or grains and subgrains from high angular resolution electron backscatter diffraction data. Together, these new techniques enable rapid measurements of individual grains and subgrains from large populations.

36 MATERIALS SCIENCE↗
DOE OSTI2021

Machine Learning Predicts the Timing and Shear Stress Evolution of Lab Earthquakes Using Active Seismic Monitoring of Fault Zone Processes

Abstract Machine learning (ML) techniques have become increasingly important in seismology and earthquake science. Lab‐based studies have used acoustic emission data to predict time‐to‐failure and stress state, and in a few cases, the same approach has been used for field data. However, the underlying physical mechanisms that allow lab earthquake prediction and seismic forecasting remain poorly resolved. Here, we address this knowledge gap by coupling active‐source seismic data, which probe asperity‐scale processes, with ML methods. We show that elastic waves passing through the lab fault zone contain information that can predict the full spectrum of labquakes from slow slip instabilities to highly aperiodic events. The ML methods utilize systematic changes in P‐wave amplitude and velocity to accurately predict the timing and shear stress during labquakes. The ML predictions improve in accuracy closer to fault failure, demonstrating that the predictive power of the ultrasonic signals improves as the fault approaches failure. Our results demonstrate that the relationship between the ultrasonic parameters and fault slip rate, and in turn, the systematically evolving real area of contact and asperity stiffness allow the gradient boosting algorithm to “learn” about the state of the fault and its proximity to failure. Broadly, our results demonstrate the utility of physics‐informed ML in forecasting the imminence of fault slip at the laboratory scale, which may have important implications for earthquake mechanics in nature.

58 GEOSCIENCES↗
DOE OSTI2021

A Cyber-Physical Anomaly Detection for Wide-Area Protection Using Machine Learning

Wide-area protection scheme (WAPS) provides system-wide protection by detecting and mitigating small and large-scale disturbances that are difficult to resolve using local protection schemes. As this protection scheme is evolving from a substation-based distributed remedial action scheme (DRAS) to the control center-based centralized RAS (CRAS), it presents severe challenges to their cybersecurity because of its heavy reliance on an insecure grid communication, and its compromise would lead to system failure. This article presents an architecture and methodology for developing a cyber-physical anomaly detection system (CPADS) that utilizes synchrophasor measurements and properties of network packets to detect data integrity and communication failure attacks on measurement and control signals in CRAS. The proposed machine leaning-based methodology applies a rules-based approach to select relevant input features, utilizes variational mode decomposition (VMD) and decision tree (DT) algorithms to develop multiple classification models, and performs final event identification using a rules-based decision logic. Here, we have evaluated the proposed methodology of CPADS using the IEEE 39 bus system for several performance measures (accuracy, recall, precision, and F-measure) in a cyber-physical testbed environment. Furthermore, our experimental results reveal that the proposed algorithm (VMD-DT) of CPADS outperforms the existing machine learning classifiers during noisy and noise-free measurements while incurring an acceptable processing overhead.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2020

Development and Validation of Quantitative Model-Based Image Reconstruction for NDE of Reinforced Steel-Lined Concrete Structures: Kal-El FY 2020

This project studied the capabilities of Model-Based Image Reconstruction (MBIR) and Machine Learning (ML) algorithms in the imaging and estimation of rebar corrosion from ultrasound signals. The application was challenging due to the introduction of a steel liner between the sensor and the concrete containing the rebar. The study focused in a synthetic specimen with a 6.35 mm thick steel liner, and a 19.05 mm diameter rebar at a depth of 44.5 mm and with four levels of corrosion; 0%, 20%, 50%, and 100% corrosion level. The corrosion was applied uniformly around the rebar, generating a ring in the perimeter of the rebar with different acoustic density. In order to generate a realistic synthetic dataset, we added random texture to the concrete of the synthetic specimens. This texture will mimic the variations encountered in real concrete specimens.We adapted our MBIR algorithm for the application and enhanced the algorithm to integrate in the physical model known specimen features, such as the steel liner thickness. The new MBIR method was able to cancel out reverberation from the steel liner and properly image the rebar. In particular, corrosion for the 50% and 100% cases were easily visible and quantitative metrics showed a slight separation for the 0% and 20% levels. The quantitative results were in agreement with the qualitative assessment. We also developed a ML XGBoost model to estimate corrosion level from the ultrasound signals. By combining the ML method with MBIR, we can pinpoint in the ultrasound signals the section that corresponds to the echoes from the rebar. The extract signals are processed by the XGBoost model for prediction. From each system scan, we obtain 15 predictions. The median prediction value is used as the final prediction. The True Acceptance Rate for the method is over 99%.

36 MATERIALS SCIENCE↗
DOE OSTI2026

Towards a self-driving trigger at the LHC: adaptive response in real time

Real-time data filtering and selection—or trigger—systems at high-throughput scientific facilities such as the experiments at the Large Hadron Collider must process extremely high-rate data streams under stringent bandwidth, latency, and storage constraints. Yet these systems are typically designed as static, hand-tuned menus of selection criteria grounded in prior knowledge and simulation. In this work, we further explore the concept of a self-driving trigger, an autonomous data-filtering framework that reallocates resources and adjusts thresholds dynamically in real-time to optimize signal efficiency, rate stability, and computational cost as instrumentation and environmental conditions evolve. We introduce a benchmark ecosystem to emulate realistic collider scenarios and demonstrate real-time optimization of a menu including canonical energy sum triggers as well as modern anomaly-detection algorithms that target non-standard event topologies using machine learning. Using simulated data streams and publicly available collision data from the Compact Muon Solenoid experiment, we demonstrate the capability to dynamically and automatically optimize trigger performance under specific cost objectives without manual retuning. Our adaptive strategy shifts trigger design from static menus with heuristic tuning to intelligent, automated, data-driven control, unlocking greater flexibility and discovery potential in future high-energy physics analyses.

Emami, Shaghayegh [Michigan U.] (ORCID:00090007589↗
DOE OSTI2025

Ca X ML: Chemistry‐informed machine learning explains mutual changes between protein conformations and calcium ions in calcium‐binding proteins using structural and topological features

Proteins' flexibility is a feature in communicating changes in cell signaling instigated by binding with secondary messengers, such as calcium ions, associated with the coordination of muscle contraction, neurotransmitter release, and gene expression. When binding with the disordered parts of a protein, calcium ions must balance their charge states with the shape of calcium-binding proteins and their versatile pool of partners depending on the circumstances they transmit. Accurately determining the ionic charges of those ions is essential for understanding their role in such processes. However, it is unclear whether the limited experimental data available can be effectively used to train models to accurately predict the charges of calcium-binding protein variants. Here, we developed a chemistry-informed, machine-learning algorithm that implements a game theoretic approach to explain the output of a machine-learning model without the prerequisite of an excessively large database for high-performance prediction of atomic charges. We used the ab initio electronic structure data representing calcium ions and the structures of the disordered segments of calcium-binding peptides with surrounding water molecules to train several explainable models. Network theory was used to extract the topological features of atomic interactions in the structurally complex data dictated by the coordination chemistry of a calcium ion, a potent indicator of its charge state in protein. Our design created a computational tool of Ca X ML, which provided a framework of explainable machine learning model to annotate ionic charges of calcium ions in calcium-binding proteins in response to the chemical changes in an environment. Our framework will provide new insights into protein design for engineering functionality based on the limited size of scientific data in a genome space.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

CSAPR2 CMAC 2.0 level c1 data.

Raw data from ARM precipitation radars must be corrected for atmospheric phenomena and instrument characteristics (e.g., attenuation, clutter) to retrieve precipitation properties. The Corrected Moments in Antenna Coordinates Version 2 (CMAC2) value-added product (VAP) is a set of algorithms and code that does such corrections, and it also retrieves precipitation quantities from the radar measurements. Similar to the X-SAPR radars at the ARM main site, C-SAPR2 radar data also needs corrections and improvements. CMAC 2.0 has been updated to work with the ARM C-SAPR 2 radar. Data and fields that have been processed to: correct for velocity aliasing, unfold and generate a cross-polarimetric phase difference that is monotonically increasing, removing impulses caused by non-uniform beam filling and phase shift on backscatter, recalculate specific differential phase using a 20-point Sobel filter on the aforementioned phase, correct for liquid path attenuation using the polarimetric signals, estimate rainfall rates at the gate using the specific attenuation. In addition, CMAC writes the data out into a community-standard format netCDF File using the CF/Radial conventions. The data are therefore compatible with new and existing National Center for Atmospheric Research (NCAR) tools such as RadXConvert for converting to a variety of popular file formats.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2020

Pulsed Thermal Tomography Nondestructive Examination of Additively Manufactured Reactor Materials (Second Annual Progress Report)

Additive manufacturing (AM) is an emerging method for cost-efficient fabrication of nuclear reactor parts. AM of metallic structures for nuclear energy applications is currently based on laser powder bed fusion (LPBF) process, which can introduce internal material flaws, such as pores and anisotropy. Integrity of AM structures needs to be evaluated nondestructively because material flaws could lead to premature failures due to exposure to high temperature, radiation and corrosive environment in a nuclear reactor. Thermal tomography (TT) provides a capability for non-destructive evaluation of sub-surface defects in arbitrary size structures. Thermal tomography is a computational method for heat diffusion-based imaging of solids, which provides 3D visualization of data from flash thermography measurements. We investigate thermal tomography imaging and nondestructive evaluation of stainless steel and nickel super alloy metallic structures produced with laser powder bed fusion (LPBF) additive manufacturing (AM) process. Metallic structures produced with LPBF contain defects, and there are limited capabilities to evaluate these structures non-destructively. Thermal tomography reconstruction of 3D apparent spatial effusivity provides information about AM structure geometry and internal material flaws. We study performance of thermal tomography in imaging of metallic structures through COMSOL computer simulations of transient heat transfer, and through reconstruction of data obtained from experimental measurements. Reconstruction of internal defects is investigated using a stainless steel 316L specimen with flat bottom hole (FBH) indentations, and Inconel 718 plate produced with laser powder bed fusion (LPBF) method, which contains imprinted hemispherical shape low density regions containing non-sintered metallic powder. The FBH’s have the same sizes as the imprinted defects in the LPBF specimens, but offer better imaging contrast. Thermal tomography reconstructions provide visualizations of internal defects, and allow for estimation of their sizes and locations. Detection sensitivity of TT is limited by noises. We investigate separation of signal from noise in thermography images using several machine learning (ML) methods, including new spatio-temporal blind source separation (STBSS) and spatio-temporal sparse dictionary learning (STSDL) methods. Performance of the ML methods is benchmarked using thermography data obtained from imaging stainless steel 316L and Inconel 718 specimens produced LPBF method with imprinted calibrated porosity defects. The ML methods are ranked by F-score and execution runtime. Finally, we investigate TT of AM stainless steel 316L specimen with imprinted internal porosity defects using relatively low-cost, small form factor infrared (IR) camera based on uncooled micro bolometer detector. Sparse coding related K-means singular value decomposition (SVD) machine learning, image processing algorithms are developed to improve quality of TT images through removal of Additive white Gaussian noise without blurring the images. Following initial qualification of an AM component for deployment in a nuclear reactor, a compact TT can also be used for in-service nondestructive evaluation (NDE). With capability to perform in-service NDE of the AM component lifecycle, TT data can be used for development of a component digital twin.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2023

Ultrafast radiographic imaging and tracking: An overview of instruments, methods, data, and applications

Ultrafast radiographic imaging and tracking (U-RadIT) use state-of-the-art ionizing particle and light sources to experimentally study sub-nanosecond transients or dynamic processes in physics, chemistry, biology, geology, materials science and other fields. These processes are fundamental to modern technologies and applications, such as nuclear fusion energy, advanced manufacturing, communication, and green transportation, which often involve one mole or more atoms and elementary particles, and thus are challenging to compute by using the first principles of quantum physics or other forward models. One of the central problems in U-RadIT is to optimize information yield through, e.g. high-luminosity X-ray and particle sources, efficient imaging and tracking detectors, novel methods to collect data, and large-bandwidth online and offline data processing, regulated by the underlying physics, statistics, and computing power. We review and highlight recent progress in: (a.) Detectors such as high-speed complementary metal-oxide semiconductor (CMOS) cameras, hybrid pixelated array detectors integrated with Timepix4 and other application-specific integrated circuits (ASICs), and digital photon detectors; (b.) U-RadIT modalities such as dynamic phase contrast imaging, dynamic diffractive imaging, and four-dimensional (4D) particle tracking; (c.) U-RadIT data and algorithms such as neural networks and machine learning, and (d.) Applications in ultrafast dynamic material science using XFELs, synchrotrons and laser-driven sources. Hardware-centric approaches to U-RadIT optimization are constrained by detector material properties, low signal-to-noise ratio, high cost and long development cycles of critical hardware components such as ASICs. Interpretation of experimental data, including comparisons with forward models, is frequently hindered by sparse measurements, model and measurement uncertainties, and noise. Alternatively, U-RadIT make increasing use of data science and machine learning algorithms, including experimental implementations of compressed sensing. Machine learning and artificial intelligence approaches, refined by physics and materials information, may also contribute significantly to data interpretation, uncertainty quantification and U-RadIT optimization.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2024

Framework of compressive sensing and data compression for 4D-STEM

Four-dimensional Scanning Transmission Electron Microscopy (4D-STEM) is a powerful technique for high-resolution and high-precision materials characterization at multiple length scales, including the characterization of beam-sensitive materials. However, the field of view of 4D-STEM is relatively small, which in absence of live processing is limited by the data size required for storage. Furthermore, the rectilinear scan approach currently employed in 4D-STEM places a resolution- and signal-dependent dose limit for the study of beam sensitive materials. Improving 4D-STEM data and dose efficiency, by keeping the data size manageable while limiting the amount of electron dose, is thus critical for broader applications. Here we introduce a general method for reconstructing 4D-STEM data with subsampling in both real and reciprocal spaces at high fidelity. The approach is first tested on the subsampled datasets created from a full 4D-STEM dataset, and then demonstrated experimentally using random scan in real-space. The same reconstruction algorithm can also be used for compression of 4D-STEM datasets, leading to a large reduction (100 times or more) in data size, while retaining the fine features of 4D-STEM imaging, for crystalline samples.

4D-STEM↗
DOE OSTI2023

Recursive Use of the Short-Time Fast Fourier Transform for Signature Analysis in Continuous Processes

Although a nuclear reactor is a hostile environment for sensing and electrical communications, the reactor core is amenable to acoustic communication. An acoustic measurement infrastructure (AMI) has been installed in the Advanced Test Reactor (ATR) to record acoustic signals that can capture its different operating regimes. AMI uses coolant pumps as continuous signal sources, coolant and structural components as transmission lines, and accelerometers to capture system motion. A recursive signal processing technique based on the short-time fast Fourier transform (STFFT) for continuous processes provides unique signatures for diagnostic and prognostic analyses from the system motion data. Here this article presents a recursive STFFT methodology that processes acoustic signals from continuous industrial processes. The article first discusses the initial STFFT use with simulated data to elucidate the basic principles necessary to understand and interpret the STFFT results from actual pump vibration data. Each repetitive use of the STFFT on pump vibration data using the results from the prior STFFT processing will generate additional complimentary time-frequency-based signatures. These signatures are generated by the coolant pumps operating under different process conditions. After each use of the STFFT, the resulting signatures provide exemplary examples of the diversity and intuitive nature of recursively using the STFFT. This article focuses on recursively using the STFFT to provide numerous complimentary and diverse signatures that will ultimately be inputs for machine learning algorithms that provide predictive data analytics. The intuitive nature of the information and signatures from recursive STFFT processing will also bring intuitive interpretation capabilities to machine learning and predictive data analytic techniques.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2024

Detection of Isotopes in Urban Source Search Gamma Spectra Using Hopfield Neural Network

Source search campaigns involve the measurement of consecutive background gamma-ray spectra obtained with a mobile detector-spectrometer, such as tellurium-doped sodium iodide NaI(Tl) scintillator detector-spectrometer. Counts are measured with NaI(Tl) detector in short acquisition intervals, such as one second. Detection of isotopes in search data is challenging because of the low signal-to-noise ratio (S/N) of the spectral lines and highly varying background due to naturally occurring radioactive material (NORM). In this work, we investigate the performance of the detection of orphan Cs-137, Am-241, and I-131 isotopes in a gamma ray data set obtained in a search survey. Isotopic detection is performed with a Hopfield Neural Network (HNN), which implements an auto-associative memory by matching incoming measurements to the closest one of its memory-stored signature patterns. HNN is trained on examples of spectra of calibrated isotopic sources with well-resolved spectral lines. During testing, HNN processes spectra with partially complete or distorted isotopic lines. The performance of the HNN detection algorithm is benchmarked with Precision, Recall, and F1 scores. Test results demonstrate the potential of HNN to detect the isotope of interest with high accuracy (F1 > 90%) for all isotopes in this study.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2021

Deep Symbolic Optimization for Electric Component Sizing in Fixed Topology Power Converters

Power converters (PC) are a major component in any current electronic hardware device. The development and design is usually guided by expert knowledge and heavily relies on human intuition and experience. The process is a very time consuming and costly activity and it is generally hard to improve upon current designs. As a first step towards autonomous PC design, we are here proposing a new framework for the sizing of components for fixed topology PCs based on given design requirements. To this end, we developed surrogate models for rapid evaluation of new topologies and adapt the deep symbolic optimization (DSO) framework to generate new topologies guided by a reinforcement learning training signal. In an empirical evaluation, we show that our DSO based approach is able to find the optimal configuration for all investigated topologies, while reducing the learning time by at least a factor of 100 compared to popular RL algorithms.

Xue, Lincoln↗
DOE OSTI2025

Semi-supervised permutation invariant particle-level anomaly detection

The development of analysis methods to distinguish potential beyond the Standard Model phenomena in a model-agnostic way can significantly enhance the discovery reach in collider experiments. However, the typical machine learning (ML) algorithms employed for this task require fixed length and ordered inputs that break the natural permutation invariance in collision events. To address this, a semi-supervised anomaly detection tool is presented that takes a variable number of particle-level inputs and leverages a signal model to encode this information into a permutation invariant, event-level representation via supervised training with a Particle Flow Network (PFN). Data events are then encoded into this representation and given as input to an autoencoder for unsupervised ANomaly deTEction on particLe flOw latent sPacE (ANTELOPE), classifying anomalous events based on a low-level and permutation invariant input modeling. Performance of the ANTELOPE architecture is evaluated on simulated samples of hadronic processes in a high energy collider experiment, showing good capability to distinguish disparate models of new physics.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2021

Binary Complex Neural Network Acceleration on FPGA

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 × 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.

Peng, Hongwu↗