Cain: Automatic Code Generation for Simultaneous Convolutional Kernels on Focal-plane Sensor-processors

Stow, Edward; Murai, Riku; Saeedi, Sajad

doi:10.48550/arxiv.2101.08715

Cited by 1 publication

(1 citation statement)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…AUKE (Debrunner et al, 2019) is a useful tool to automatically generate convolution kernel codes on the PPA. Stow et al (2021) is another compiler targeting the SCAMP-5 vision system developed by Stow et al Furthermore, Martel et al trained neural networks to learn pixelwise exposures for HDR imaging and video compression (Martel et al, 2020).…”

Section: Related Studymentioning

confidence: 99%

On-sensor binarized CNN inference with dynamic model swapping in pixel processor arrays

et al. 2022

View full text Add to dashboard Cite

Many types of Convolutional Neural Network (CNN) models and training methods have been proposed in recent years aiming to provide efficiency for embedded and edge devices with limited computation and memory resources. The wide variety of architectures makes this a complex task that has to balance generality with efficiency. Among the most interesting camera-sensor architectures are Pixel Processor Arrays (PPAs). This study presents two methods that are useful for embedded CNNs in general but particularly suitable for PPAs. The first is for training purely binarized CNNs, the second is for deploying larger models with a model swapping paradigm that loads model components dynamically. Specifically, this study trains and implements networks with batch normalization and adaptive threshold for binary activations. Then, we convert batch normalization and binary activations into a bias matrix which can be parallelly implemented by an add/sub operation. For dynamic model swapping, we propose to decompose applications that are beyond the capacity of a PPA into sub-tasks that can be solved by tree networks that can be loaded dynamically as needed. We demonstrate our approaches to various tasks including classification, localization, and coarse segmentation on a highly resource constrained PPA sensor-processor.

show abstract

Section: Related Studymentioning

confidence: 99%