摘要
一位新闻记者兼机器人与机器学习的新闻编辑每日新闻-关于人工智能的新研究结果已经发表。据NewsRx编辑的《来自吉隆坡的新闻》Mal Aysia报道，这项研究称：“全景立体视频以其沉浸感和立体效果为观众带来了全新的视觉体验。”我们的新闻记者从大学学院的研究中得到一句话：“在全景立体视频中，人脸是一个重要的元素。”然而，全景立体视频中的人脸图像具有不同程度的变形。这给人脸识别带来了新的挑战。为此，本文提出了一种适用于全景立体视觉的人脸识别模型DCM2Net（可变形卷积MobileFaceNet）。该模型主要是在特征融合过程中整合通道间的特征信息，在网络的深层层次对通道间的信息进行重新分配，充分利用通道间的信息进行特征提取。本文还建立了全景立体视频直播系统，利用DCM2Net模型对全景立体视频中的人脸进行识别，并将识别结果显示在视频中。
Abstract
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-New study results on artificial intell igence have been published. According to news originating from Kuala Lumpur, Mal aysia, by NewsRx editors, the research stated, "The panoramic stereo video has b rought a new visual experience for the audience with its immersion and stereo ef fect." Our news reporters obtained a quote from the research from University College: " In panoramic stereo video, the face is an important element. However, the face i mage in panoramic stereo video has varying degrees of deformation. This brings n ew challenges to face recognition. Therefore, this paper proposes a face recogni tion model DCM2Net (Deformable Convolution MobileFaceNet) for panoramic stereo v ideo. The model mainly integrates the feature information between channels durin g feature fusion, redistributes the information between channels in the deeper p art of the network, and fully uses the information between different channels fo r feature extraction. This paper also built a panoramic stereo video live system , using the DCM2Net model to recognize the face in panoramic stereo video, and t he recognition results are displayed in the video."
NETL
NSTL