Diverse sample generation with multi-branch conditional generative adversarial network for remote sensing objects detection

Efficient-Lightweight YOLO: Improving Small Object Detection in YOLO for Aerial Images

The most significant technical challenges of current aerial image object-detection tasks are the extremely low accuracy for detecting small objects that are densely distributed within a scene and the lack of semantic information. Moreover, existing detectors with large parameter scales are unsuitable for aerial image object-detection scenarios oriented toward low-end GPUs. To address this technical challenge, we propose efficient-lightweight You Only Look Once (EL-YOLO), an innovative model that overcomes the limitations of existing detectors and low-end GPU orientation. EL-YOLO surpasses the baseline models in three key areas. Firstly, we design and scrutinize three model architectures to intensify the model's focus on small objects and identify the most effective network structure. Secondly, we design efficient spatial pyramid pooling (ESPP) to augment the representation of small-object features in aerial images. Lastly, we introduce the alpha-complete intersection over union (α-CIoU) loss function to tackle the imbalance between positive and negative samples in aerial images. Our proposed EL-YOLO method demonstrates a strong generalization and robustness for the small-object detection problem in aerial images. The experimental results show that, with the model parameters maintained below 10 M while the input image size was unified at 640 × 640 pixels, the AP_S of the EL-YOLOv5 reached 10.8% and 10.7% and enhanced the APs by 1.9% and 2.2% compared to YOLOv5 on two challenging aerial image datasets, DIOR and VisDrone, respectively.

Computational ghost imaging based on a conditional generation countermeasure network under a low sampling rate

In this paper, an end-to-end depth neural network based on a conditional generative adversarial network for computational ghost imaging (CGANCGI) is proposed to restore clear object images with high quality at a sub-Nyquist sampling rate. The 2D light signal collected by a CMOS camera and the gray image of the original measured object are used as the input of the network model; then, the CGANCGI network is trained, and the measured object image is recovered directly from the 2D light signal. Experiments have verified that the proposed method only needs 1/10 of traditional deep learning samples to achieve fast image restoration with high-quality, and its peak signal-to-noise ratio and structural similarity are, respectively, four to six times and five to seven times higher than those of the original image, which prove that our method has practical application prospects in ghost imaging under low sampling rates.