CA-Cut: Crop-Aligned Cutout for Data Augmentation to Learn More Robust Under-Canopy Navigation

TL;DR

This paper introduces CA-Cut, a novel data augmentation technique that masks regions around crop rows in images to improve the robustness and accuracy of under-canopy navigation models in complex agricultural environments.

Contribution

The paper proposes CA-Cut, a crop-aligned masking augmentation method that enhances model robustness by simulating occlusions and focusing on high-level contextual features.

Findings

CA-Cut significantly improves semantic keypoint prediction accuracy.

Biasing masks toward crop rows enhances model generalization.

Up to 36.9% reduction in prediction error achieved.

Abstract

State-of-the-art visual under-canopy navigation methods are designed with deep learning-based perception models to distinguish traversable space from crop rows. While these models have demonstrated successful performance, they require large amounts of training data to ensure reliability in real-world field deployment. However, data collection is costly, demanding significant human resources for in-field sampling and annotation. To address this challenge, various data augmentation techniques are commonly employed during model training, such as color jittering, Gaussian blur, and horizontal flip, to diversify training data and enhance model robustness. In this paper, we hypothesize that utilizing only these augmentation techniques may lead to suboptimal performance, particularly in complex under-canopy environments with frequent occlusions, debris, and non-uniform spacing of crops. Instead, we propose a novel augmentation method, so-called Crop-Aligned Cutout (CA-Cut) which masks random regions out in input images that are spatially distributed around crop rows on the sides to encourage trained models to capture high-level contextual features even when fine-grained information is obstructed. Our extensive experiments with a public cornfield dataset demonstrate that masking-based augmentations are effective for simulating occlusions and significantly improving robustness in semantic keypoint predictions for visual navigation. In particular, we show that biasing the mask distribution toward crop rows in CA-Cut is critical for enhancing both prediction accuracy and generalizability across diverse environments achieving up to a 36.9% reduction in prediction error. In addition, we conduct ablation studies to determine the number of masks, the size of each mask, and the spatial distribution of masks to maximize overall performance.