ERNet Family: Hardware-Oriented CNN Models for Computational Imaging Using Block-Based Inference

10/13/2019
by   Chao-Tsung Huang, et al.
0

Convolutional neural networks (CNNs) demand huge DRAM bandwidth for computational imaging tasks, and block-based processing has recently been applied to greatly reduce the bandwidth. However, the induced additional computation for feature recomputing or the large SRAM for feature reusing will degrade the performance or even forbid the usage of state-of-the-art models. In this paper, we address these issues by considering the overheads and hardware constraints in advance when constructing CNNs. We investigate a novel model family—ERNet—which includes temporary layer expansion as another means for increasing model capacity. We analyze three ERNet variants in terms of hardware requirement and introduce a hardware-aware model optimization procedure. Evaluations on Full HD and 4K UHD applications will be given to show the effectiveness in terms of image quality, pixel throughput, and SRAM usage. The results also show that, for block-based inference, ERNet can outperform the state-of-the-art FFDNet and EDSR-baseline models for image denoising and super-resolution respectively.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
10/13/2019

eCNN: A Block-Based and Highly-Parallel CNN Accelerator for Edge Inference

Convolutional neural networks (CNNs) have recently demonstrated superior...
research
05/02/2022

BSRA: Block-based Super Resolution Accelerator with Hardware Efficient Pixel Attention

Increasingly, convolution neural network (CNN) based super resolution mo...
research
03/17/2021

Collapsible Linear Blocks for Super-Efficient Super Resolution

With the advent of smart devices that support 4K and 8K resolution, Sing...
research
03/05/2019

Towards Design Space Exploration and Optimization of Fast Algorithms for Convolutional Neural Networks (CNNs) on FPGAs

Convolutional Neural Networks (CNNs) have gained widespread popularity i...
research
08/05/2019

Architecture-aware Network Pruning for Vision Quality Applications

Convolutional neural network (CNN) delivers impressive achievements in c...
research
04/19/2021

RingCNN: Exploiting Algebraically-Sparse Ring Tensors for Energy-Efficient CNN-Based Computational Imaging

In the era of artificial intelligence, convolutional neural networks (CN...
research
01/21/2021

GhostSR: Learning Ghost Features for Efficient Image Super-Resolution

Modern single image super-resolution (SISR) system based on convolutiona...

Please sign up or login with your details

Forgot password? Click here to reset