{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/driver-gaze-zone-estimation-using","title":"Driver Gaze Zone Estimation using Convolutional Neural Networks: A General Framework and Ablative Analysis","arxiv_id":"1802.02690","date":"2018-02-08","proceeding":null,"authors":["Sourabh Vora","Akshay Rangesh","Mohan M. Trivedi"],"abstract":"Driver gaze has been shown to be an excellent surrogate for driver attention\nin intelligent vehicles. With the recent surge of highly autonomous vehicles,\ndriver gaze can be useful for determining the handoff time to a human driver.\nWhile there has been significant improvement in personalized driver gaze zone\nestimation systems, a generalized system which is invariant to different\nsubjects, perspectives and scales is still lacking. We take a step towards this\ngeneralized system using Convolutional Neural Networks (CNNs). We finetune 4\npopular CNN architectures for this task, and provide extensive comparisons of\ntheir outputs. We additionally experiment with different input image patches,\nand also examine how image size affects performance. For training and testing\nthe networks, we collect a large naturalistic driving dataset comprising of 11\nlong drives, driven by 10 subjects in two different cars. Our best performing\nmodel achieves an accuracy of 95.18% during cross-subject testing,\noutperforming current state of the art techniques for this task. Finally, we\nevaluate our best performing model on the publicly available Columbia Gaze\nDataset comprising of images from 56 subjects with varying head pose and gaze\ndirections. Without any training, our model successfully encodes the different\ngaze directions on this diverse dataset, demonstrating good generalization\ncapabilities.","url_abs":"http://arxiv.org/abs/1802.02690v2","url_pdf":"http://arxiv.org/pdf/1802.02690v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[],"tasks":[{"task_slug":"autonomous-vehicles","task_name":"Autonomous Vehicles"}],"methods":[],"datasets_introduced":[{"slug":"lisa-gaze-dataset","name":"LISA Gaze Dataset","full_name":""}],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1802.02690","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}