The paper introduces HCF-Net, a deep learning method designed to improve the performance of infrared small object detection. The method addresses the challenges of small object loss and low background distinctiveness in infrared images through three key modules: Parallelized Patch-Aware Attention (PPA), Dimension-Aware Selective Integration (DASI), and Multi-Dilated Channel Refiner (MDCR). PPA uses a multi-branch feature extraction strategy to capture multi-scale features, DASI enhances skip connections by adaptively selecting and fusing high and low-dimensional features, and MDCR captures spatial features across different receptive field ranges using depth-separable convolutional layers. Experimental results on the SIRST dataset show that HCF-Net outperforms other traditional and deep learning models, achieving high Intersection over Union (IoU) and normalized IoU (nIoU) scores. The paper also discusses related work in traditional and deep learning methods for infrared small object detection and provides detailed descriptions of the proposed modules, loss function design, and implementation details.The paper introduces HCF-Net, a deep learning method designed to improve the performance of infrared small object detection. The method addresses the challenges of small object loss and low background distinctiveness in infrared images through three key modules: Parallelized Patch-Aware Attention (PPA), Dimension-Aware Selective Integration (DASI), and Multi-Dilated Channel Refiner (MDCR). PPA uses a multi-branch feature extraction strategy to capture multi-scale features, DASI enhances skip connections by adaptively selecting and fusing high and low-dimensional features, and MDCR captures spatial features across different receptive field ranges using depth-separable convolutional layers. Experimental results on the SIRST dataset show that HCF-Net outperforms other traditional and deep learning models, achieving high Intersection over Union (IoU) and normalized IoU (nIoU) scores. The paper also discusses related work in traditional and deep learning methods for infrared small object detection and provides detailed descriptions of the proposed modules, loss function design, and implementation details.