Local and Global Fusion Network For Learned Image Compression

The paper presents a framework to identify whether a document image is perturbed with glare. Glare identification on documents is particularly challenging because of predominantly white background of documents and dearth of training dataset. The paper addresses the dataset bottleneck by introducing a glare synthesis framework to render a million training images. The training model consists of a global deep neural network supplemented by localized feature extraction. To our best knowledge, this is one of the first works towards classifying document image for presence of glare. Experiments on real glare dataset showcase benefits of combined global and local learning and also outperform recent glare segmentation models adapted for the classification task.