Remote Sensing Image Mosaicking: Achievements and Challenges Page 1 8 0274-6638/19©2019IEEE
IEEE GEOSCIENCE AND REMOTE SENSING MAGAZINE DECEMBER 2019 Image …

Plain text has become a prevalent interface for text-to-image synthesis. However, its limited
customization options hinder users from accurately describing desired outputs. For example …

We present a novel technique to automatically colorize grayscale images that combines
both global priors and local image features. Based on Convolutional Neural Networks, our …

We develop a fully automatic image colorization system. Our approach leverages recent
advances in deep networks, exploiting both low-level and semantic representations. As …

We propose a new technique for visual attribute transfer across images that may have very
different appearance but have perceptually similar semantic structure. By visual attribute …

We present the Colorization Transformer, a novel approach for diverse high fidelity image
colorization based on self-attention. Given a grayscale image, the colorization proceeds in …

We propose the first deep learning approach for exemplar-based local colorization. Given a
reference color image, our convolutional neural network directly maps a grayscale image to …

Automatic quantification and classification of leukocytes in microscopic images are of
paramount importance in the perspective of disease identification, its progress and drugs …

The colorization of grayscale images is an ill-posed problem, with multiple correct solutions.
In this paper, we propose an adversarial learning colorization approach coupled with …

This paper presents the first end-to-end network for exemplar-based video colorization. The
main challenge is to achieve temporal consistency while remaining faithful to the reference …