# Adaptive Pixel-wise Structured Sparse Network for Efficient CNNs

Source: [Adaptive Pixel-wise Structured Sparse Network for Efficient CNNs](<https://www.sota2.com/research/paper/68fa698d1e209131cea228e3>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-25T23:02:46.333Z

Authors: Chen Tang, Wenyu Sun, Zhuqing Yuan, Yongpan Liu

Published: 2020-10-21T15:47:13.000Z

## Abstract

To accelerate deep CNN models, this paper proposes a novel spatially adaptive framework that can dynamically generate pixel-wise sparsity according to the input image. The sparse scheme is pixel-wise refined, regional adaptive under a unified importance map, which makes it friendly to hardware implementation. A sparse controlling method is further presented to enable online adjustment for applications with different precision/latency requirements. The sparse model is applicable to a wide range of vision tasks. Experimental results show that this method efficiently improve the computing efficiency for both image classification using ResNet-18 and super resolution using SRResNet. On image classification task, our method can save 30%-70% MACs with a slightly drop in top-1 and top-5 accuracy. On super resolution task, our method can reduce more than 90% MACs while only causing around 0.1 dB and 0.01 decreasing in PSNR and SSIM. Hardware validation is also included.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2010.11083>)

[PDF](<https://arxiv.org/pdf/2010.11083.pdf>)