 We present a novel aircraft detection algorithm for remote sensing images, RSIs, with low signal-to-noise ratio, LSNYOLO. Our approach combines a SWIN transformer and convolutional neural network, CNN, to extract multi-scale global and local RSI information. We then introduce an effective feature enhancement, EFE, blocked to further improve the detection accuracy. Finally, we use a novel loss function to optimize the detection accuracy. Experimental results show that our LSNYOLO outperforms other state-of-the-art methods on RSIs with low SNR. This article was authored by Ruizinho, Shiyoung Ji, Shurekai Jung, and others.