Following Gaze in Video

Following the gaze of people inside videos is an important signal for understanding people and their actions. In this paper, we present an approach for following gaze in video by predicting where a person (in the video) is looking even when the object is in a different frame. We collect VideoGaze, a...

Full description

Bibliographic Details
Main Authors: Recasens Continente, Adria, Vondrick, Carl Martin, Khosla, Aditya, Torralba, Antonio
Other Authors: Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Format: Article
Language:English
Published: Institute of Electrical and Electronics Engineers 2019
Online Access:https://hdl.handle.net/1721.1/122778
_version_ 1826190795776983040
author Recasens Continente, Adria
Vondrick, Carl Martin
Khosla, Aditya
Torralba, Antonio
author2 Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
author_facet Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Recasens Continente, Adria
Vondrick, Carl Martin
Khosla, Aditya
Torralba, Antonio
author_sort Recasens Continente, Adria
collection MIT
description Following the gaze of people inside videos is an important signal for understanding people and their actions. In this paper, we present an approach for following gaze in video by predicting where a person (in the video) is looking even when the object is in a different frame. We collect VideoGaze, a new dataset which we use as a benchmark to both train and evaluate models. Given one frame with a person in it, our model estimates a density for gaze location in every frame and the probability that the person is looking in that particular frame. A key aspect of our approach is an end-to-end model that jointly estimates: saliency, gaze pose, and geometric relationships between views while only using gaze as supervision. Visualizations suggest that the model learns to internally solve these intermediate tasks automatically without additional supervision. Experiments show that our approach follows gaze in video better than existing approaches, enabling a richer understanding of human activities in video. Keywords: Motion pictures, Head, Three-dimensional displays, Predictive models, Geometry, Semantics, gaze tracking, learning (artificial intelligence), video signal processing
first_indexed 2024-09-23T08:45:48Z
format Article
id mit-1721.1/122778
institution Massachusetts Institute of Technology
language English
last_indexed 2024-09-23T08:45:48Z
publishDate 2019
publisher Institute of Electrical and Electronics Engineers
record_format dspace
spelling mit-1721.1/1227782022-09-23T14:22:58Z Following Gaze in Video Recasens Continente, Adria Vondrick, Carl Martin Khosla, Aditya Torralba, Antonio Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Following the gaze of people inside videos is an important signal for understanding people and their actions. In this paper, we present an approach for following gaze in video by predicting where a person (in the video) is looking even when the object is in a different frame. We collect VideoGaze, a new dataset which we use as a benchmark to both train and evaluate models. Given one frame with a person in it, our model estimates a density for gaze location in every frame and the probability that the person is looking in that particular frame. A key aspect of our approach is an end-to-end model that jointly estimates: saliency, gaze pose, and geometric relationships between views while only using gaze as supervision. Visualizations suggest that the model learns to internally solve these intermediate tasks automatically without additional supervision. Experiments show that our approach follows gaze in video better than existing approaches, enabling a richer understanding of human activities in video. Keywords: Motion pictures, Head, Three-dimensional displays, Predictive models, Geometry, Semantics, gaze tracking, learning (artificial intelligence), video signal processing 2019-11-06T20:19:42Z 2019-11-06T20:19:42Z 2017-12 2019-07-11T16:25:39Z Article http://purl.org/eprint/type/ConferencePaper 2380-7504 https://hdl.handle.net/1721.1/122778 Recasens Continente, Adria et al. "Following Gaze in Video," 2017 IEEE International Conference on Computer Vision (ICCV), October 2017, Venice, Italy, Institute of Electrical and Electronics Engineers, December 2017 ©IEEE en http://dx.doi.org/10.1109/iccv.2017.160 2017 IEEE International Conference on Computer Vision (ICCV) Creative Commons Attribution-Noncommercial-Share Alike http://creativecommons.org/licenses/by-nc-sa/4.0/ application/pdf Institute of Electrical and Electronics Engineers MIT web domain
spellingShingle Recasens Continente, Adria
Vondrick, Carl Martin
Khosla, Aditya
Torralba, Antonio
Following Gaze in Video
title Following Gaze in Video
title_full Following Gaze in Video
title_fullStr Following Gaze in Video
title_full_unstemmed Following Gaze in Video
title_short Following Gaze in Video
title_sort following gaze in video
url https://hdl.handle.net/1721.1/122778
work_keys_str_mv AT recasenscontinenteadria followinggazeinvideo
AT vondrickcarlmartin followinggazeinvideo
AT khoslaaditya followinggazeinvideo
AT torralbaantonio followinggazeinvideo