Salient region detection in the task of visual question answering

9th International Multidisciplinary Scientific and Research Conference "Modern Issues in Science and Technology" Workshop "Advanced Technologies in Aerospace, Mechanical and Automation Engineering"

自然科学;工业技术

Favorskaya, Margarita^1 ; Andreev, Vladimir^1 ; Popov, Aleksei^1

Reshetnev Siberian State University of Science and Technology, 31 Krasnoyarsky Rabochy ave., Krasnoyarsk

660037, Russia^1

关键词: Conventional approach;    Convolutional neural network;    High probability;    Human abilities;    Neural network application;    Question Answering;    Salient region detections;    Segmentation results;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/450/5/052017/pdf
DOI  :  10.1088/1757-899X/450/5/052017

来源: IOP

【 摘 要 】

Salient region detection in Visual Question Answering (VQA) is an attempt to simulate a human ability to quickly perceive a scene by selectively looking on image fragments instead of processing a whole scene. The conventional approach deals with a neural network application. However, the Convolutional Neural Networks (CNNs) have many disadvantages compared with traditional methods for salient region detection. We modified the basic algorithm of salient region detection for VQA task by selecting such image fragments, which have a high probability to be included in a questionnaire. The experiments have been conducted on images from MS-COCO dataset and provided good segmentation results.