2021/09/27 by Nairit Bandyopadhyay, Sébastien Riou, Bandyopadhyay, Nairit +3
Computer Science · Medicine · #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Gaze Tracking and Assistive Technology #Retinal Imaging and Analysis #Visual Attention and Saliency Detection
paper · pdf · doi:10.48550/arxiv.2109.12801
openalex publication_date 2021/09/27 · openalex created_date 2022/07/25 · openalex updated_date 2026/07/28
With the increase in computation power and the development of new\nstate-of-the-art deep learning algorithms, appearance-based gaze estimation is\nbecoming more and more popular. It is believed to work well with curated\nlaboratory data sets, however it faces several challenges when deployed in real\nworld scenario. One such challenge is to estimate the gaze of a person about\nwhich the Deep Learning model trained for gaze estimation has no knowledge\nabout. To analyse the performance in such scenarios we have tried to simulate a\ncalibration mechanism. In this work we use the MPIIGaze data set. We trained a\nmulti modal convolutional neural network and analysed its performance with and\nwithout calibration and this evaluation provides clear insights on how\ncalibration improved the performance of the Deep Learning model in estimating\ngaze in the wild.\n