2Net) that utilizes an adaptive hybrid attention and multi-scale perception strategy to incrementally fuse multi-scale features. Specifically, we design an adaptive feature fusion (AFF) module, which mines deep complex features and shallow positional information through graph channel attention and pixel attention, respectively, and fully integrates the complementary information between the features so as to fuse the features effectively. Then the multi-scale perception (MSP) module is designed to learn multi-scale information through three-branch convolutional blocks of different scales to more accurately localize camouflaged objects of different scales and guide the next layer of feature fusion. Extensive experiments on three large benchmark datasets show the superior performance of our approach. Our source code and results can be found at https://github.com/akuan1234/AF2Net.">

Adaptive Feature Fusion Network for Camouflaged Object Detection (original) (raw)

IEEE Account

Purchase Details

Profile Information

Need Help?

A not-for-profit organization, IEEE is the world's largest technical professional organization dedicated to advancing technology for the benefit of humanity.
© Copyright 2026 IEEE - All rights reserved. Use of this web site signifies your agreement to the terms and conditions.