[feed] Atom [feed] RSS 1.0 [feed] RSS 2.0

Parekh, V and Subramanian, R and Roy, D and Jawahar, CV (2018) An EEG-based image annotation system. Communications in Computer and Information Science, 841. pp. 303-313.

[img] Text
An EEG-based Image Annotation System.pdf
Restricted to Repository staff only

Download (1055Kb) | Request a copy

Abstract

The success of deep learning in computer vision has greatly increased the need for annotated image datasets. We propose an EEG (Electroencephalogram)-based image annotation system. While humans can recognize objects in 20-200 milliseconds, the need to manually label images results in a low annotation throughput. Our system employs brain signals captured via a consumer EEG device to achieve an annotation rate of up to 10 images per second. We exploit the P300 event-related potential (ERP) signature to identify target images during a rapid serial visual presentation (RSVP) task. We further perform unsupervised outlier removal to achieve an F1-score of 0.88 on the test set. The proposed system does not depend on category-speci�c EEG signatures enabling the annotation of any new image category without any model pre-training.

Item Type: Article
Subjects: Neurodegenerative Disorders
Neuro-Oncological Disorders
Neurocognitive Processes
Neuronal Development and Regeneration
Informatics and Imaging
Genetics and Molecular Biology
Depositing User: Dr. D.D. Lal
Date Deposited: 10 Aug 2018 06:55
Last Modified: 10 Aug 2018 06:55
URI: http://nbrc.sciencecentral.in/id/eprint/429

Actions (login required)

View Item View Item