Adaptive Pixel-wise Structured Sparse Network for Efficient CNNs

2020-10-21 · Chen Tang, Wenyu Sun, Zhuqing Yuan, Yongpan Liu

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.

📄 PDF Abstract BibTeX arXiv:2010.11083

Code (0)

등록된 구현이 없습니다.

Tasks

Denoising General Classification image-classification Image Classification Image Denoising SSIM Style Transfer Super-Resolution

Structured Priors for Sparse-Representation-Based Hyperspectral Image Classification

2014-01-16 · Xiaoxia Sun, Qing Qu, Nasser M. Nasrabadi, Trac. D. Tran

Pixel-wise classification, where each pixel is assigned to a predefined class, is one of the most important procedures in hyperspectral image (HSI) analysis. By representing a test pixel as a linear combination of a smal…

Classification General Classification Hyperspectral Image Classification image-classification+1

A Versatile Framework for Designing Group-Sparse Adversarial Attacks

2025-10-18 · Alireza Heshmati, Saman Soleimani Roudi, Sajjad Amini, Shahrokh Ghaemmaghami 외 arxiv

Existing adversarial attacks often neglect perturbation sparsity, limiting their ability to model structural changes and to explain how deep neural networks (DNNs) process meaningful input patterns. We propose ATOS (Atta…

PSGR: Pixel-wise Sparse Graph Reasoning for COVID-19 Pneumonia Segmentation in CT Images

2021-08-09 · Haozhe Jia, Haoteng Tang, Guixiang Ma, Weidong Cai 외

Automated and accurate segmentation of the infected regions in computed tomography (CT) images is critical for the prediction of the pathological stage and treatment response of COVID-19. Several deep convolutional neura…

Computed Tomography (CT)graph construction Information Retrieval Retrieval+1

Learning Propagation for Arbitrarily-structured Data

2019-09-25 · ICCV 2019 10 · Sifei Liu, Xueting Li, Varun Jampani, Shalini De Mello 외

Processing an input signal that contains arbitrary structures, e.g., superpixels and point clouds, remains a big challenge in computer vision. Linear diffusion, an effective model for image processing, has been recently …

Point Cloud Segmentation Segmentation Semantic Segmentation Superpixels

Propagating Confidences through CNNs for Sparse Data Regression

2018-05-30 · Abdelrahman Eldesokey, Michael Felsberg, Fahad Shahbaz Khan

In most computer vision applications, convolutional neural networks (CNNs) operate on dense image data generated by ordinary cameras. Designing CNNs for sparse and irregularly spaced input data is still an open problem w…

Autonomous Driving Depth Completion regression