Deep learning models are found to be vulnerable adversarial examples, as wrong predictions can caused by small perturbation in input for deep models. Most of the existing works image generation try achieve attacks most models, while few them make efforts on guaranteeing perceptual quality examples. High examples matter many applications, especially privacy preserving. In this work, we develop a...

The Image Difference Captioning (IDC) task aims to describe the visual differences between two similar images with natural language. major challenges of this lie in aspects: 1) fine-grained that require learning stronger vision and language association 2) high-cost manual annotations leads limited supervised data. To address these challenges, we propose a new modeling framework following pre-tr...

Imaging and evaporation of atoms in the field ion microscope (FIM) has been modelled by using finite difference methods to calculate the voltage distribution around a tip and hence the electric field strength experienced by individual atoms. Atoms are evaporated based on field strength using a number of different mathematical models which yield broadly similar results. The tip shapes and simula...

An image difference structure is created that include of image regularization, feature derivation, and feature grouping. Based on this structure image difference measure is created by selecting exact implementations for each of the steps. An image difference prediction is then calculated using image difference features (IDFs) that are derived from the images. The input images are then regulariz...