Deep forgery detection technologies are crucial for image and video recognition tasks, with their performance heavily reliant on the features extracted from both real and fake images. However, most existing methods primarily focus on spatial domain features, which limits their accuracy. To address this limitation, we propose an adaptive dual-domain feature representation method for enhanced deep forgery detection. Specifically, an adaptive region dynamic convolution module is established to efficiently extract facial features from the spatial domain. Then, we introduce an adaptive frequency dynamic filter to capture effective frequency domain features. By fusing both spatial and frequency domain features, our approach significantly improves the accuracy of classifying real and fake facial images. Finally, experimental results on three real-world datasets validate the effectiveness of our dual-domain feature representation method, which substantially improves classification precision.

In order to meet the intelligent demand of waste sorting and the centralized management of classification equipment, an intelligent waste sorting system based on cloud edge collaboration was designed and developed. The system uses Raspberry Pi and Alibaba Cloud HaaS100 development board as the edge end master equipment, and Alibaba Cloud IOT platform as the cloud platform. The system selects 245 categories of waste sorting dataset provided by Baidu PaddlePaddle, and combines data enhancement and cloud edge collaboration technology to train, test and deploy three lightweight waste sorting algorithms. The experimental results show that data augmentation can improve the Top-1 accuracy of the algorithm model on the test set by about 1%, and cloud edge collaboration can achieve a Top-1 accuracy of 98.80% for Raspberry Pi in inference, which can meet practical usage needs.