Localizing Semantic Patches for Accelerating Image Classification

06/07/2022
by   Chuanguang Yang, et al.
0

Existing works often focus on reducing the architecture redundancy for accelerating image classification but ignore the spatial redundancy of the input image. This paper proposes an efficient image classification pipeline to solve this problem. We first pinpoint task-aware regions over the input image by a lightweight patch proposal network called AnchorNet. We then feed these localized semantic patches with much smaller spatial redundancy into a general classification network. Unlike the popular design of deep CNN, we aim to carefully design the Receptive Field of AnchorNet without intermediate convolutional paddings. This ensures the exact mapping from a high-level spatial location to the specific input image patch. The contribution of each patch is interpretable. Moreover, AnchorNet is compatible with any downstream architecture. Experimental results on ImageNet show that our method outperforms SOTA dynamic inference methods with fewer inference costs. Our code is available at https://github.com/winycg/AnchorNet.

READ FULL TEXT

page 3

page 6

research
01/31/2020

Localizing Multi-scale Semantic Patches for Image Classification

Deep convolutional neural networks (CNN) always non-linearly aggregate t...
research
07/29/2021

PPT Fusion: Pyramid Patch Transformerfor a Case Study in Image Fusion

The Transformer architecture has achieved rapiddevelopment in recent yea...
research
05/07/2021

Adaptive Focus for Efficient Video Recognition

In this paper, we explore the spatial redundancy in video recognition wi...
research
08/19/2022

Accelerating Vision Transformer Training via a Patch Sampling Schedule

We introduce the notion of a Patch Sampling Schedule (PSS), that varies ...
research
10/11/2020

Glance and Focus: a Dynamic Approach to Reducing Spatial Redundancy in Image Classification

The accuracy of deep convolutional neural networks (CNNs) generally impr...
research
11/11/2020

FPGA: Fast Patch-Free Global Learning Framework for Fully End-to-End Hyperspectral Image Classification

Deep learning techniques have provided significant improvements in hyper...
research
01/09/2022

Glance and Focus Networks for Dynamic Visual Recognition

Spatial redundancy widely exists in visual recognition tasks, i.e., disc...

Please sign up or login with your details

Forgot password? Click here to reset