Semantic Representation for Visual Reasoning

Know about the book -

Semantic Representation for Visual Reasoning is written by Wenfeng Zheng and published by . It's available with International Standard Book Number or ISBN identification 1954318030 (ISBN 10) and 9781954318038 (ISBN 13).

Visual Question Answering (VQA) combines natural language processing with digital image processing. The general process for solving a VQA problem is to take the image and the corresponding question as input and finally get the answer. The problems which are similar to VQA require more interdependent inference steps to solve.The research is mainly divided into the non-deep learning model and deep learning model. Most non-deep learning models are based on Bayesian theory. Those non-deep learning model performed poorly. In the field of visual reasoning, the achievement of deep learning significantly improved the accuracy of results. With the deepening of the research on deep learning, its research on the VQA field is becoming more and more mature. Image features are mostly used as the input to get answers. However, the image features are too redundant to learn accurate characterizations within a limited complexity and time. While in the process of human reasoning, abstract description of an image is usually to avoid irrelevant details. Inspired by this, a higher-level representation named semantic representation is introduced. In this book, the authors propose a dataset that weakens image processing and natural language processing. The idea of the Gram matrix is transferred here to build a relationship to represent the related information well. The model using semantic representation as input verifies that more accurate results can be obtained by introducing a high-level semantic representation.