Efficiency in Real-time Webcam Gaze Tracking

Amogh Gudi; Xin li; Jan van Gemert

doi:10.1007/978-3-030-66415-2_34

Efficiency in Real-time Webcam Gaze Tracking

Pattern Recognition and Bioinformatics

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

10 Citations (Scopus)

78 Downloads (Pure)

Abstract

Efficiency and ease of use are essential for practical applications of camera based eye/gaze-tracking. Gaze tracking involves estimating where a person is looking on a screen based on face images from a computer-facing camera. In this paper we investigate two complementary forms of efficiency in gaze tracking: 1. The computational efficiency of the system which is dominated by the inference speed of a CNN predicting gaze-vectors; 2. The usability efficiency which is determined by the tediousness of the mandatory calibration of the gaze-vector to a computer screen. To do so, we evaluate the computational speed/accuracy trade-off for the CNN and the calibration effort/accuracy trade-off for screen calibration. For the CNN, we evaluate the full face, two-eyes, and single eye input. For screen calibration, we measure the number of calibration points needed and evaluate three types of calibration: 1. pure geometry, 2. pure machine learning, and 3. hybrid geometric regression. Results suggest that a single eye input and geometric regression calibration achieve the best trade-off.

Original language	English
Title of host publication	Computer Vision – ECCV 2020 Workshops
Subtitle of host publication	Proceedings
Editors	Adrien Bartoli, Andrea Fusiello
Place of Publication	Cham
Publisher	Springer
Chapter	34
Pages	529 - 543
Number of pages	15
Edition	1
ISBN (Electronic)	978-3-030-66415-2
ISBN (Print)	978-3-030-66414-5
DOIs	https://doi.org/10.1007/978-3-030-66415-2_34
Publication status	Published - 2020
Event	European Conference on Computer Vision (ECCV) 2020 Workshop on Eye Gaze in AR, VR, and in the Wild (OpenEyes) - Glassgow, United Kingdom Duration: 23 Aug 2020 → 28 Aug 2020 https://openeyes-workshop.github.io/

Publication series

Name	Part of the Lecture Notes in Computer Science book series (LNCS, volume 12535) Also part of the Image Processing, Computer Vision, Pattern Recognition, and Graphics book sub series (LNIP, volume 12535)
Publisher	Springer
Volume	12535
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Workshop

Workshop	European Conference on Computer Vision (ECCV) 2020 Workshop on Eye Gaze in AR, VR, and in the Wild (OpenEyes)
Abbreviated title	ECCVW 2020; OpenEyes 2020
Country/Territory	United Kingdom
City	Glassgow
Period	23/08/20 → 28/08/20
Internet address	https://openeyes-workshop.github.io/

Access to Document

10.1007/978-3-030-66415-2_34

2009.01270-1Accepted author manuscript, 1.03 MB

Cite this

Gudi, A., li, X., & van Gemert, J. (2020). Efficiency in Real-time Webcam Gaze Tracking. In A. Bartoli, & A. Fusiello (Eds.), Computer Vision – ECCV 2020 Workshops: Proceedings (1 ed., pp. 529 - 543). (Part of the Lecture Notes in Computer Science book series (LNCS, volume 12535) Also part of the Image Processing, Computer Vision, Pattern Recognition, and Graphics book sub series (LNIP, volume 12535); Vol. 12535). Springer. https://doi.org/10.1007/978-3-030-66415-2_34

Gudi, Amogh ; li, Xin ; van Gemert, Jan. / Efficiency in Real-time Webcam Gaze Tracking. Computer Vision – ECCV 2020 Workshops: Proceedings. editor / Adrien Bartoli ; Andrea Fusiello. 1. ed. Cham : Springer, 2020. pp. 529 - 543 (Part of the Lecture Notes in Computer Science book series (LNCS, volume 12535) Also part of the Image Processing, Computer Vision, Pattern Recognition, and Graphics book sub series (LNIP, volume 12535)).

@inproceedings{0724c10199974206abd387d5e4a65f99,

title = "Efficiency in Real-time Webcam Gaze Tracking",

abstract = "Efficiency and ease of use are essential for practical applications of camera based eye/gaze-tracking. Gaze tracking involves estimating where a person is looking on a screen based on face images from a computer-facing camera. In this paper we investigate two complementary forms of efficiency in gaze tracking: 1. The computational efficiency of the system which is dominated by the inference speed of a CNN predicting gaze-vectors; 2. The usability efficiency which is determined by the tediousness of the mandatory calibration of the gaze-vector to a computer screen. To do so, we evaluate the computational speed/accuracy trade-off for the CNN and the calibration effort/accuracy trade-off for screen calibration. For the CNN, we evaluate the full face, two-eyes, and single eye input. For screen calibration, we measure the number of calibration points needed and evaluate three types of calibration: 1. pure geometry, 2. pure machine learning, and 3. hybrid geometric regression. Results suggest that a single eye input and geometric regression calibration achieve the best trade-off.",

author = "Amogh Gudi and Xin li and {van Gemert}, Jan",

year = "2020",

doi = "10.1007/978-3-030-66415-2_34",

language = "English",

isbn = "978-3-030-66414-5",

series = "Part of the Lecture Notes in Computer Science book series (LNCS, volume 12535) Also part of the Image Processing, Computer Vision, Pattern Recognition, and Graphics book sub series (LNIP, volume 12535)",

publisher = "Springer",

pages = "529 -- 543",

editor = "Adrien Bartoli and Andrea Fusiello",

booktitle = "Computer Vision – ECCV 2020 Workshops",

edition = "1",

note = "European Conference on Computer Vision (ECCV) 2020 Workshop on Eye Gaze in AR, VR, and in the Wild (OpenEyes), ECCVW 2020; OpenEyes 2020 ; Conference date: 23-08-2020 Through 28-08-2020",

url = "https://openeyes-workshop.github.io/",

}

Gudi, A, li, X & van Gemert, J 2020, Efficiency in Real-time Webcam Gaze Tracking. in A Bartoli & A Fusiello (eds), Computer Vision – ECCV 2020 Workshops: Proceedings. 1 edn, Part of the Lecture Notes in Computer Science book series (LNCS, volume 12535) Also part of the Image Processing, Computer Vision, Pattern Recognition, and Graphics book sub series (LNIP, volume 12535), vol. 12535, Springer, Cham, pp. 529 - 543, European Conference on Computer Vision (ECCV) 2020 Workshop on Eye Gaze in AR, VR, and in the Wild (OpenEyes), Glassgow, United Kingdom, 23/08/20. https://doi.org/10.1007/978-3-030-66415-2_34

Efficiency in Real-time Webcam Gaze Tracking. / Gudi, Amogh; li, Xin; van Gemert, Jan.
Computer Vision – ECCV 2020 Workshops: Proceedings. ed. / Adrien Bartoli; Andrea Fusiello. 1. ed. Cham: Springer, 2020. p. 529 - 543 (Part of the Lecture Notes in Computer Science book series (LNCS, volume 12535) Also part of the Image Processing, Computer Vision, Pattern Recognition, and Graphics book sub series (LNIP, volume 12535); Vol. 12535).

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

TY - GEN

T1 - Efficiency in Real-time Webcam Gaze Tracking

AU - Gudi, Amogh

AU - li, Xin

AU - van Gemert, Jan

PY - 2020

Y1 - 2020

N2 - Efficiency and ease of use are essential for practical applications of camera based eye/gaze-tracking. Gaze tracking involves estimating where a person is looking on a screen based on face images from a computer-facing camera. In this paper we investigate two complementary forms of efficiency in gaze tracking: 1. The computational efficiency of the system which is dominated by the inference speed of a CNN predicting gaze-vectors; 2. The usability efficiency which is determined by the tediousness of the mandatory calibration of the gaze-vector to a computer screen. To do so, we evaluate the computational speed/accuracy trade-off for the CNN and the calibration effort/accuracy trade-off for screen calibration. For the CNN, we evaluate the full face, two-eyes, and single eye input. For screen calibration, we measure the number of calibration points needed and evaluate three types of calibration: 1. pure geometry, 2. pure machine learning, and 3. hybrid geometric regression. Results suggest that a single eye input and geometric regression calibration achieve the best trade-off.

AB - Efficiency and ease of use are essential for practical applications of camera based eye/gaze-tracking. Gaze tracking involves estimating where a person is looking on a screen based on face images from a computer-facing camera. In this paper we investigate two complementary forms of efficiency in gaze tracking: 1. The computational efficiency of the system which is dominated by the inference speed of a CNN predicting gaze-vectors; 2. The usability efficiency which is determined by the tediousness of the mandatory calibration of the gaze-vector to a computer screen. To do so, we evaluate the computational speed/accuracy trade-off for the CNN and the calibration effort/accuracy trade-off for screen calibration. For the CNN, we evaluate the full face, two-eyes, and single eye input. For screen calibration, we measure the number of calibration points needed and evaluate three types of calibration: 1. pure geometry, 2. pure machine learning, and 3. hybrid geometric regression. Results suggest that a single eye input and geometric regression calibration achieve the best trade-off.

UR - http://www.scopus.com/inward/record.url?scp=85101327825&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-66415-2_34

DO - 10.1007/978-3-030-66415-2_34

M3 - Conference contribution

SN - 978-3-030-66414-5

T3 - Part of the Lecture Notes in Computer Science book series (LNCS, volume 12535) Also part of the Image Processing, Computer Vision, Pattern Recognition, and Graphics book sub series (LNIP, volume 12535)

SP - 529

EP - 543

BT - Computer Vision – ECCV 2020 Workshops

A2 - Bartoli, Adrien

A2 - Fusiello, Andrea

PB - Springer

CY - Cham

T2 - European Conference on Computer Vision (ECCV) 2020 Workshop on Eye Gaze in AR, VR, and in the Wild (OpenEyes)

Y2 - 23 August 2020 through 28 August 2020

ER -

Gudi A, li X, van Gemert J. Efficiency in Real-time Webcam Gaze Tracking. In Bartoli A, Fusiello A, editors, Computer Vision – ECCV 2020 Workshops: Proceedings. 1 ed. Cham: Springer. 2020. p. 529 - 543. (Part of the Lecture Notes in Computer Science book series (LNCS, volume 12535) Also part of the Image Processing, Computer Vision, Pattern Recognition, and Graphics book sub series (LNIP, volume 12535)). doi: 10.1007/978-3-030-66415-2_34

Efficiency in Real-time Webcam Gaze Tracking

Abstract

Publication series

Workshop

Access to Document

Other files and links

Fingerprint

Less Machine (=) More Vision: Approaches towards Practical and Efficient Machine Vision with Applications in Face Analysis

Best Paper Award

Cite this

Efficiency in Real-time Webcam Gaze Tracking

Abstract

Publication series

Workshop

Access to Document

Other files and links

Fingerprint

Research output

Less Machine (=) More Vision: Approaches towards Practical and Efficient Machine Vision with Applications in Face Analysis

Prizes

Best Paper Award

Cite this