TdAttenMix: Top-Down Attention Guided Mixup

TL;DR

TdAttenMix introduces a novel data augmentation method that uses human gaze-guided top-down attention to improve image mixing, leading to better performance across multiple benchmarks.

Contribution

It is the first to incorporate human gaze-driven top-down attention into mixup data augmentation, balancing high-level and low-level cues for more label-consistent image mixing.

Findings

Outperforms state-of-the-art mixup methods on eight benchmarks.

Introduces a human gaze-based metric for image-label consistency.

Demonstrates improved training effectiveness with attention-guided patch selection.

Abstract

CutMix is a data augmentation strategy that cuts and pastes image patches to mixup training data. Existing methods pick either random or salient areas which are often inconsistent to labels, thus misguiding the training model. By our knowledge, we integrate human gaze to guide cutmix for the first time. Since human attention is driven by both high-level recognition and low-level clues, we propose a controllable Top-down Attention Guided Module to obtain a general artificial attention which balances top-down and bottom-up attention. The proposed TdATttenMix then picks the patches and adjust the label mixing ratio that focuses on regions relevant to the current label. Experimental results demonstrate that our TdAttenMix outperforms existing state-of-the-art mixup methods across eight different benchmarks. Additionally, we introduce a new metric based on the human gaze and use this metric to investigate the issue of image-label inconsistency. Project page: \url{https://github.com/morning12138/TdAttenMix}