This paper introduces the Poly Kernel Inception Network (PKINet) for remote sensing object detection, addressing challenges such as large variations in object scales and diverse contextual information. PKINet employs multi-scale convolution kernels without dilation to extract features of varying scales and capture local context. A Context Anchor Attention (CAA) module is introduced to capture long-range contextual information. These components work together to enhance performance on four remote sensing benchmarks: DOTA-v1.0, DOTA-v1.5, HRSC2016, and DIOR-R. PKINet outperforms existing methods in terms of accuracy and efficiency, achieving state-of-the-art results with a lightweight design. The network uses inception-style depth-wise convolutions and a CAA mechanism to adaptively extract features with both local and global contextual information. Experimental results show that PKINet achieves high performance on various remote sensing datasets, demonstrating its effectiveness in handling object detection challenges in remote sensing images.This paper introduces the Poly Kernel Inception Network (PKINet) for remote sensing object detection, addressing challenges such as large variations in object scales and diverse contextual information. PKINet employs multi-scale convolution kernels without dilation to extract features of varying scales and capture local context. A Context Anchor Attention (CAA) module is introduced to capture long-range contextual information. These components work together to enhance performance on four remote sensing benchmarks: DOTA-v1.0, DOTA-v1.5, HRSC2016, and DIOR-R. PKINet outperforms existing methods in terms of accuracy and efficiency, achieving state-of-the-art results with a lightweight design. The network uses inception-style depth-wise convolutions and a CAA mechanism to adaptively extract features with both local and global contextual information. Experimental results show that PKINet achieves high performance on various remote sensing datasets, demonstrating its effectiveness in handling object detection challenges in remote sensing images.