{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,6,19]],"date-time":"2025-06-19T05:06:34Z","timestamp":1750309594653,"version":"3.41.0"},"reference-count":45,"publisher":"Association for Computing Machinery (ACM)","issue":"2","license":[{"start":{"date-parts":[[2025,5,26]],"date-time":"2025-05-26T00:00:00Z","timestamp":1748217600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/www.acm.org\/publications\/policies\/copyright_policy#Background"}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["Proc. ACM Comput. Graph. Interact. Tech."],"published-print":{"date-parts":[[2025,6]]},"abstract":"<jats:p>The task of pupil localization is an imperative step for any gaze estimation and eye tracking systems. Due to the challenges induced by diverse factors like illumination, reflections and noise, researchers have proposed several classical and learning based approaches addressing this problem. Even though learning based methods have shown superior accuracy compared to classical methods, they require extensive annotations for training and are prone to poor cross-dataset performance on new unseen scenarios. In this work, we formulated pupil center estimation as an unpaired image-to-image translation task and introduced a novel loss function to the existing CycleGAN framework to achieve robust and accurate pupil localization using self-supervised learning paradigm. Further, we evaluated our method on four publicly available datasets, Labelled pupils in the wild (LPW), OpenEDS, \u015awirski, and NVGaze, and reported the state-of-the-art pupil localization performance. \u015awirski, and NVGaze are used for cross-data evaluation to understand the generalizability of the proposed method. The proposed method achieved an improvement of around 16% on LPW, 21% on OpenEDS, 5% on \u015awirski, and 14% on NVGaze datasets in terms of pupil detection rate over the conventional methods.<\/jats:p>","DOI":"10.1145\/3729416","type":"journal-article","created":{"date-parts":[[2025,5,27]],"date-time":"2025-05-27T05:51:27Z","timestamp":1748325087000},"page":"1-16","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["S3PL : Self-Supervised Segmentation for Robust Pupil Localization"],"prefix":"10.1145","volume":"8","author":[{"ORCID":"https:\/\/orcid.org\/0009-0005-4736-1617","authenticated-orcid":false,"given":"Purma","family":"VishnuVardhan","sequence":"first","affiliation":[{"name":"Mercedes Benz Research and Development India, Bangalore, India"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0003-8353-6387","authenticated-orcid":false,"given":"Vignan","family":"Gummadi","sequence":"additional","affiliation":[{"name":"Mercedes Benz Research and Development India, Bangalore, India"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0003-6314-8555","authenticated-orcid":false,"given":"Ashwini Kundranda","family":"Poovaiah","sequence":"additional","affiliation":[{"name":"Mercedes Benz Research and Development India, Bengaluru, Karanataka, India"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7039-6763","authenticated-orcid":false,"given":"Lrd","family":"Murthy","sequence":"additional","affiliation":[{"name":"Mercedes Benz Research and Development India, Bangalore, India"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"320","published-online":{"date-parts":[[2025,5,26]]},"reference":[{"key":"e_1_3_1_2_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.eswa.2021.116004"},{"key":"e_1_3_1_3_1","first-page":"214","volume-title":"International conference on machine learning","author":"Arjovsky Martin","year":"2017","unstructured":"Martin Arjovsky, Soumith Chintala, and L\u00e9on Bottou. 2017. Wasserstein generative adversarial networks. In International conference on machine learning. PMLR, 214\u2013223."},{"key":"e_1_3_1_4_1","article-title":"Loss functions in the era of semantic segmentation: A survey and outlook","author":"Azad Reza","year":"2023","unstructured":"Reza Azad, Moein Heidary, Kadir Yilmaz, Michael H\u00fcttemann, Sanaz Karimijafarbigloo, Yuli Wu, Anke Schmeink, and Dorit Merhof. 2023. Loss functions in the era of semantic segmentation: A survey and outlook. arXiv preprint arXiv:https:\/\/arXiv.org\/abs\/2312.05391 (2023).","journal-title":"arXiv preprint arXiv:https:\/\/arXiv.org\/abs\/2312.05391"},{"key":"e_1_3_1_5_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.eswa.2020.114431"},{"key":"e_1_3_1_6_1","doi-asserted-by":"publisher","DOI":"10.1109\/ITSC.2015.268"},{"key":"e_1_3_1_7_1","doi-asserted-by":"publisher","DOI":"10.1145\/3448018.3458009"},{"key":"e_1_3_1_8_1","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2021.3068094"},{"key":"e_1_3_1_9_1","doi-asserted-by":"publisher","DOI":"10.1145\/3649902.3653351"},{"key":"e_1_3_1_10_1","doi-asserted-by":"publisher","DOI":"10.1145\/3314111.3319914"},{"key":"e_1_3_1_11_1","doi-asserted-by":"publisher","DOI":"10.1109\/ICCVW.2019.00541"},{"key":"e_1_3_1_12_1","doi-asserted-by":"publisher","DOI":"10.1145\/3204493.3204559"},{"key":"e_1_3_1_13_1","doi-asserted-by":"publisher","DOI":"10.1109\/ISMAR52148.2021.00053"},{"key":"e_1_3_1_14_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-23192-1_4"},{"key":"e_1_3_1_15_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-29888-3_27"},{"key":"e_1_3_1_16_1","article-title":"Pupilnet: Convolutional neural networks for robust pupil detection","author":"Fuhl Wolfgang","year":"2016","unstructured":"Wolfgang Fuhl, Thiago Santini, Gjergji Kasneci, and Enkelejda Kasneci. 2016a. Pupilnet: Convolutional neural networks for robust pupil detection. arXiv preprint arXiv:https:\/\/arXiv.org\/abs\/1601.04902 (2016).","journal-title":"arXiv preprint arXiv:https:\/\/arXiv.org\/abs\/1601.04902"},{"key":"e_1_3_1_17_1","doi-asserted-by":"publisher","DOI":"10.1145\/2857491.2857505"},{"key":"e_1_3_1_18_1","article-title":"Openeds: Open eye dataset","author":"Garbin Stephan\u00a0J","year":"2019","unstructured":"Stephan\u00a0J Garbin, Yiru Shen, Immo Schuetz, Robert Cavin, Gregory Hughes, and Sachin\u00a0S Talathi. 2019. Openeds: Open eye dataset. arXiv preprint arXiv:https:\/\/arXiv.org\/abs\/1905.03702 (2019).","journal-title":"arXiv preprint arXiv:https:\/\/arXiv.org\/abs\/1905.03702"},{"key":"e_1_3_1_19_1","article-title":"Escaf: Pupil centre localization algorithm with candidate filtering","author":"George Anjith","year":"2018","unstructured":"Anjith George and Aurobinda Routray. 2018. Escaf: Pupil centre localization algorithm with candidate filtering. arXiv preprint arXiv:https:\/\/arXiv.org\/abs\/1807.10520 (2018).","journal-title":"arXiv preprint arXiv:https:\/\/arXiv.org\/abs\/1807.10520"},{"key":"e_1_3_1_20_1","article-title":"Generative adversarial nets","volume":"27","author":"Goodfellow Ian","year":"2014","unstructured":"Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. 2014. Generative adversarial nets. Advances in neural information processing systems 27 (2014).","journal-title":"Advances in neural information processing systems"},{"key":"e_1_3_1_21_1","volume-title":"Remote, non-contact gaze estimation with minimal subject cooperation","author":"Guestrin Elias\u00a0Daniel","year":"2010","unstructured":"Elias\u00a0Daniel Guestrin. 2010. Remote, non-contact gaze estimation with minimal subject cooperation. Vol.\u00a0395. University of Toronto."},{"key":"e_1_3_1_22_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2000.855815"},{"key":"e_1_3_1_23_1","doi-asserted-by":"publisher","DOI":"10.1109\/CIBCB48159.2020.9277638"},{"key":"e_1_3_1_24_1","doi-asserted-by":"publisher","DOI":"10.3389\/fneng.2015.00004"},{"key":"e_1_3_1_25_1","doi-asserted-by":"publisher","DOI":"10.1145\/3290605.3300780"},{"key":"e_1_3_1_26_1","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.324"},{"key":"e_1_3_1_27_1","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.304"},{"key":"e_1_3_1_28_1","article-title":"Revisiting CycleGAN for semi-supervised segmentation","author":"Mondal Arnab\u00a0Kumar","year":"2019","unstructured":"Arnab\u00a0Kumar Mondal, Aniket Agarwal, Jose Dolz, and Christian Desrosiers. 2019. Revisiting CycleGAN for semi-supervised segmentation. arXiv preprint arXiv:https:\/\/arXiv.org\/abs\/1908.11569 (2019).","journal-title":"arXiv preprint arXiv:https:\/\/arXiv.org\/abs\/1908.11569"},{"key":"e_1_3_1_29_1","doi-asserted-by":"publisher","DOI":"10.1109\/AERO47225.2020.9172480"},{"key":"e_1_3_1_30_1","article-title":"Semi-supervised domain adaptation with CycleGAN guided by a downstream task loss","author":"M\u00fctze Annika","year":"2022","unstructured":"Annika M\u00fctze, Matthias Rottmann, and Hanno Gottschalk. 2022. Semi-supervised domain adaptation with CycleGAN guided by a downstream task loss. arXiv preprint arXiv:https:\/\/arXiv.org\/abs\/2208.08815 (2022).","journal-title":"arXiv preprint arXiv:https:\/\/arXiv.org\/abs\/2208.08815"},{"key":"e_1_3_1_31_1","doi-asserted-by":"crossref","unstructured":"Graham Naylor Thomas Koelewijn Adriana\u00a0A Zekveld and Sophia\u00a0E Kramer. 2018. The application of pupillometry in hearing science to assess listening effort. 2331216518799437\u00a0pages.","DOI":"10.1177\/2331216518799437"},{"key":"e_1_3_1_32_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.treng.2020.100008"},{"key":"e_1_3_1_33_1","doi-asserted-by":"publisher","DOI":"10.1109\/TMI.2024.3453492"},{"key":"e_1_3_1_34_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-67389-9_44"},{"key":"e_1_3_1_35_1","doi-asserted-by":"publisher","DOI":"10.1145\/3208031.3208032"},{"key":"e_1_3_1_36_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.cviu.2018.02.002"},{"key":"e_1_3_1_37_1","doi-asserted-by":"publisher","DOI":"10.1145\/633292.633452"},{"key":"e_1_3_1_38_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-67558-9_28"},{"key":"e_1_3_1_39_1","doi-asserted-by":"publisher","DOI":"10.1145\/2807442.2807445"},{"key":"e_1_3_1_40_1","doi-asserted-by":"publisher","DOI":"10.1145\/2168556.2168585"},{"key":"e_1_3_1_41_1","doi-asserted-by":"publisher","DOI":"10.1145\/2857491.2857520"},{"key":"e_1_3_1_42_1","doi-asserted-by":"publisher","DOI":"10.3233\/ICA-180584"},{"key":"e_1_3_1_43_1","doi-asserted-by":"publisher","DOI":"10.1109\/ISIMP.2004.1434171"},{"key":"e_1_3_1_44_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.jneumeth.2019.05.016"},{"key":"e_1_3_1_45_1","article-title":"Correlation Maximized Structural Similarity Loss for Semantic Segmentation","volume":"1910","author":"Zhao Shuai","year":"2019","unstructured":"Shuai Zhao, Boxi Wu, Wenqing Chu, Yao Hu, and Deng Cai. 2019. Correlation Maximized Structural Similarity Loss for Semantic Segmentation. ArXiv abs\/1910.08711 (2019).","journal-title":"ArXiv"},{"key":"e_1_3_1_46_1","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.244"}],"container-title":["Proceedings of the ACM on Computer Graphics and Interactive Techniques"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3729416","content-type":"unspecified","content-version":"vor","intended-application":"text-mining"}],"deposited":{"date-parts":[[2025,6,19]],"date-time":"2025-06-19T01:56:57Z","timestamp":1750298217000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3729416"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,5,26]]},"references-count":45,"journal-issue":{"issue":"2","published-print":{"date-parts":[[2025,6]]}},"alternative-id":["10.1145\/3729416"],"URL":"https:\/\/doi.org\/10.1145\/3729416","relation":{},"ISSN":["2577-6193"],"issn-type":[{"type":"electronic","value":"2577-6193"}],"subject":[],"published":{"date-parts":[[2025,5,26]]},"assertion":[{"value":"2025-05-26","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}