Due to the limited memory of Graphical Processing Units (GPUs), most current 3D GAN models are trained on low-resolution medical images, these models either cannot scale to high-resolution or are ...

In this work, we introduced a new methodology to analyze the quality of high-resolution fingerprint images containing sets of fingerprint pores. Our approach takes into account the spatial ...

Abstract: Deep convolutional neural networks (CNNs) have been successfully applied on no-reference image quality assessment (NR-IQA) with respect to human perception. Most of these methods deal with ...

Abstract: The retrieval of encrypted images in cloud computing is a research hotspot at present. However, the existing schemes have the problem of low image retrieval accuracy since it is difficult to ...