Analysis of Visual Question Answering Algorithms with attention model
It provides a critical examination of existing VQA methods, which could benefit applications like assistive technology for visually impaired people, security surveillance, and chatbots, but is incremental as it reviews rather than proposes new techniques.
This paper reviews visual question answering (VQA) algorithms that use attention models to process images and natural language, focusing on methods for text semantics, object identification, and answer classification with co-attention approaches.
Visual question answering (VQA) usesimage processing algorithms to process the image and natural language processing methods to understand and answer the question. VQA is helpful to a visually impaired person, can be used for the security surveillance system and online chatbots that learn from the web. It uses NLP methods to learn the semantic of the question and to derive the textual features. Computer vision techniques are used for generating image representation in such a way that they can identify the objects about which question is asked. The Attention model tries to mimic the human behavior of giving attention to a different region of an image according to our understanding of its context. This paper critically examines and reviews methods of VQA algorithm such as generation of semantics of text, identification of objects and answer classification techniques that use the co-attention approach.