For Full-Text PDF, please login, if you are a member of IEICE,|
or go to Pay Per View on menu list, if you are a nonmember of IEICE.
Realtime Microsaccade Detection with Convolutional Neural Network
Joshua EMOTO Yutaka HIRATA
D - Abstracts of IEICE TRANSACTIONS on Information and Systems (Japanese Edition)
Publication Date: 2018/02/01
Online ISSN: 1881-0225
Type of Manuscript: PAPER
eye movement, microsaccade, covert attention, deep learning,
Full Text(in Japanese): PDF(722.2KB)
>>Buy this Article
Microsaccade (MSC) is a rapid and miniature eye movement, which recently has been considered to reflect covert attention. There are many potential applications of MSC where estimation of attentional states is demanded. For such purposes, a real time, robust and highly accurate MSC detection system is required. Presently, we propose a novel MSC detection system, using a convolutional neural network and deep learning. We demonstrate that the proposed system runs in real time on a general PC, and outperforms representative conventional MSC detection methods in terms of accuracy by using real eye movement data recorded in human subjects.