Due to limited features and dense object layouts, object detection in UAV images is challenging. Given that existing feature fusion methods have not fully explored the relationship between fine- and coarse-grained features, direct feature fusion can result in poor correlation between them, hindering the representative capability of fine-grained semantic information. To alleviate this issue, we introduce a method of Coarse-fine Feature Interaction Alignment (CFIA), which enhances the correlation between coarse-grained and fine-grained features across multi-scale feature maps through their interactive alignment. Firstly, we present the Wavelet-based High-frequency Preserving Down-sampling (WHPD), utilizing wavelet transform to extract high-frequency information to enhance object boundaries, minimizing crucial fine-grained information loss. Secondly, we propose the Feature Refinement and Interaction Alignment Strategy (FRIAS), which achieves feature interaction alignment by establishing the association of feature maps between coarse-grained and fine-grained features. This enhances the representative capability of feature maps at various scales for detecting small objects. Extensive experiments on the VisDrone, CARPK, and Drone-vs-Bird datasets have demonstrated the effectiveness of the CFIA method, which is highly competitive with state-of-the-art methods. The code is available at https://github.com/b-yanchao/CFIA.git.

Current feature fusion strategies often fail to adequately account for the influence of activation intensity across different scales on small object features, which impedes the effective detection of small objects. To address this limitation, we propose the Region-Adaptive Feature Disentanglement and Enhancement (RAFDE) strategy, which improves both downsampling and feature fusion by leveraging activation intensity variations at multiple scales. First, we introduce the Boundary Transitional Region-enhanced Downsampling (BTRD) module, which enhances boundary transitional regions containing both strongly and weakly activated features, thereby mitigating the loss of crucial boundary information for small objects. Second, we present the Regional-Adaptive Feature Fusion (RAFF) module, which adaptively disentangles and fuses co-activated and uni-activated regions from adjacent levels into the current level, effectively reducing the risk of small objects being overwhelmed. Extensive experiments on several public datasets demonstrate that the RAFDE strategy is highly effective and outperforms state-of-the-art methods. The code is available at https://github.com/b-yanchao/RAFDE.git.

In recent years, Out-of-Distribution (o.o.d) Robustness has garnered increasing attention in Deep Learning, and shape-biased Convolutional Neural Networks (CNNs) are believed to exhibit higher robustness, attributed to the inherent shape-based decision rule of human cognition. In this work, we delve deeper into the intricate relationship between shape/texture information and o.o.d robustness by leveraging a carefully curated "Category-Balanced ImageNet" dataset. We find that shape information is not always superior in distinguishing distinct categories and shape-biased model is not always superior across various o.o.d scenarios. Motivated by these insightful findings, we design a novel method named Shape-Texture Adaptive Recombination (STAR) to achieve higher o.o.d robustness. A category-balanced dataset is firstly used to pretrain a debiased backbone and three specialized heads, each adept at robustly extracting shape, texture, and debiased features. Subsequently, an instance-adaptive recombination head is trained to adaptively adjust the contributions of these distinctive features for each given instance. Through comprehensive experiments, our proposed method achieves state-of-the-art o.o.d robustness across various scenarios such as image corruptions, adversarial attacks, style shifts, and dataset shifts, demonstrating its effectiveness.