摘要
Robotics&Machine Learning Daily News Daily News的新闻记者兼新闻编辑-根据NewsRx记者从华盛顿特区发回的新闻报道，Inve Ntors Boyle，Kevin(加利福尼亚州旧金山)；Konrad，Robert(加利福尼亚州旧金山)；Padmanaban，Nitish(加利福尼亚州门洛帕克，美国)于2022年11月15日提交的专利申请本专利申请没有受让人。记者从发明人提供的背景信息中获得了以下引文：“眼睛跟踪系统捕获眼睛的图像，以便确定用户的3D注视，或该注视在表面上或平面上的2D投影，这是通过计算机视觉将眼睛图像分割成不同部分来完成的，即瞳孔、巩膜、虹膜、眼睑、眼角等，其特征是作为参数导出的N，所述参数可以用于基于校准数据计算用户的注视，或者生成用于相同目的的眼睛模型，或者眼睛图像S直接输入神经网络或其他机器学习方法，该方法基于标记的眼睛图像数据库直接从图像中推断分割和/或用户视线。从传统的计算机视觉方法提取的参数也可以与机器学习方法一起使用，无论是否使用眼睛图像，在所有情况下，相对于对比度、亮度、灵敏度等的图像质量，以及提取眼睛的有限元特征或直接从图像推断凝视所需的计算量对于凝视估计的鲁棒性和质量是首要的。移动系统旨在室内和室外操作，在不受控制和可变的光照条件下。从眼睛图像中提取信息的复杂性，特别是关键的瞳孔位置，要求用于该任务的计算机视觉算法具有很高的复杂性，而对这些图像的环境影响的鲁棒性是眼跟踪系统面临的主要挑战。
Abstract
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News – According to news reporting originatin g from Washington, D.C., by NewsRx journalists, a patent application by the inve ntors Boyle, Kevin (San Francisco, CA, US); Konrad, Robert (San Francisco, CA, U S); Padmanaban, Nitish (Menlo Park, CA, US), filed on November 15, 2022, was mad e available online on May 16, 2024. No assignee for this patent application has been made. Reporters obtained the following quote from the background information supplied by the inventors: “Eye-tracking systems capture images of the eyes in order to d etermine the 3D gaze of the user, or a 2D projection of that gaze onto a surface or plane, such as a screen or typical viewing distance. This is done either thr ough a computer vision segmentation of the image of the eye into various parts, i.e. pupil, sclera, iris, eye lids, canthus, etc., the features of which are the n exported as parameters that can be used to calculate the user’s gaze based on calibration data or generate an eye model for the same purpose, or the eye image s are fed directly into a neural network or other machine learning approach that infers the segmentation and/or user’s gaze directly from the images based on a database of labeled eye images. The parameters extracted from a traditional comp uter vision approach can also be used with a machine learning approach, with or without the images of the eyes, which may also be scaled to various lower resolu tions. In all cases, the quality of the images, with respect to contrast, lighti ng, sensitivity, etc., and the amount of computation required to extract the fea tures of the eye or infer the gaze directly from the images is of first importan ce to the robustness and quality of the gaze estimate. This is especially true i n a head-mounted, mobile system intended to operate both indoors and outdoors, i n uncontrolled and variable lighting conditions. The complexity of extracting in formation from the eye images, especially the crucial pupil position, requires h igh complexity in the computer vision algorithms used for the task, and robustne ss to environmental effects on those images is the main challenge remaining for eye-tracking systems.”
NETL
NSTL