An energy-efficient embedded implementation for target recognition in SAR imageries

Renz, Megan; Wu, Qing

doi:10.1109/ssci.2017.8285211

Cited by 3 publications

(2 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Images are grayscale and roughly 128x128 in dimension, and each testing class contains around 300 images. To map SAR classification to the RANC ecosystem, we first preprocess the imagery in a similar approach taken by Renz and Wu [31] as illustrated in Figure 9. MSTAR dataset images vary in size from class to class, so we normalize all images to 128x128 first, then take a 64x64 image chip from the center of this resized image.…”

Section: B Convolutionmentioning

confidence: 99%

“…Renz and Wu deployed a DCNN on IBM's TrueNorth to classify SAR images and achieved an accuracy of 95.6% [31] with the expense of utilizing 4042 of the 4096 cores on a single TrueNorth chip. Because of the insufficient information on their network architecture and unavailability of IBM's Energy-Efficient Deep Neuromorphic Network (EEDN) learning framework [33], we were unable to create the exact architecture.…”

Section: B Convolutionmentioning

confidence: 99%

See 1 more Smart Citation

RANC: Reconfigurable Architecture for Neuromorphic Computing

Mack

Purdy

Rockowitz

et al. 2021

IEEE Trans. Comput.-Aided Des. Integr. Circuits Syst.

View full text Add to dashboard Cite

Neuromorphic architectures have been introduced as platforms for energy efficient spiking neural network execution. The massive parallelism offered by these architectures has also triggered interest from nonmachine learning application domains. In order to lift the barriers to entry for hardware designers and application developers we present RANC: a Reconfigurable Architecture for Neuromorphic Computing, an opensource highly flexible ecosystem that enables rapid experimentation with neuromorphic architectures in both software via C++ simulation and hardware via FPGA emulation. We present the utility of the RANC ecosystem by showing its ability to recreate behavior of the IBM's TrueNorth and validate with direct comparison to IBM's Compass simulation environment and published literature. RANC allows optimizing architectures based on application insights as well as prototyping future neuromorphic architectures that can support new classes of applications entirely. We demonstrate the highly parameterized and configurable nature of RANC by studying the impact of architectural changes on improving application mapping efficiency with quantitative analysis based on Alveo U250 FPGA. We present post routing resource usage and throughput analysis across implementations of Synthetic Aperture Radar classification and Vector Matrix Multiplication applications, and demonstrate a neuromorphic architecture that scales to emulating 259K distinct neurons and 73.3M distinct synapses.

show abstract

Section: B Convolutionmentioning

confidence: 99%

Section: B Convolutionmentioning

confidence: 99%

RANC: Reconfigurable Architecture for Neuromorphic Computing

Mack

Purdy

Rockowitz

et al. 2021

IEEE Trans. Comput.-Aided Des. Integr. Circuits Syst.

View full text Add to dashboard Cite

show abstract

Pseudo-Labels for Supervised Learning on Dynamic Vision Sensor Data, Applied to Object Detection Under Ego-Motion

Chen

2018

2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)

View full text Add to dashboard Cite

In recent years, dynamic vision sensors (DVS), also known as event-based cameras or neuromorphic sensors, have seen increased use due to various advantages over conventional frame-based cameras. Using principles inspired by the retina, its high temporal resolution overcomes motion blurring, its high dynamic range overcomes extreme illumination conditions and its low power consumption makes it ideal for embedded systems on platforms such as drones and self-driving cars. However, event-based data sets are scarce and labels are even rarer for tasks such as object detection. We transferred discriminative knowledge from a state-of-the-art frame-based convolutional neural network (CNN) to the event-based modality via intermediate pseudo-labels, which are used as targets for supervised learning. We show, for the first time, event-based car detection under ego-motion in a real environment at 100 frames per second with a test average precision of 40.3% relative to our annotated ground truth. The event-based car detector handles motion blur and poor illumination conditions despite not explicitly trained to do so, and even complements frame-based CNN detectors, suggesting that it has learnt generalized visual representations.

show abstract