IEEE IV 2018“智能车辆中的平行视觉”研讨会

IEEE IV 2018“智能车辆中的平行视觉”研讨会征稿通知

近年来，计算机视觉领域（包括目标检测、分割与跟踪、场景理解等）取得的研究进展，极大地促进了智能车辆的发展。为了建立可靠的车载视觉感知系统，需要收集大规模的标注图像数据，来训练、测试、理解和优化视觉感知模型。然而，从真实世界收集和标注大规模多样化的图像数据集，费时费力且容易出错。于是，我国学者将视觉计算研究从单一的物理世界扩展到虚实结合的平行世界（物理世界+虚拟世界），提出了平行视觉理论与方法。平行视觉由人工场景、计算实验、平行执行三个步骤组成，通过生成和利用虚拟图像，构建更健壮的智能视觉系统，正吸引着计算机视觉和智能车辆研究者越来越多的关注。

本研讨会旨在引发学术界对平行视觉的学术讨论，加快其在智能车辆中的应用。征稿的主题范围包括但不限于：

●基于计算机图形学和虚拟现实的交通场景仿真

●驾驶模拟器与虚拟图像生成

●与虚拟图像相关的生成式模型（包括生成式对抗网络、变分自编码器等）

●基于虚拟图像的神经网络与学习系统

●基于虚拟图像的智能视觉计算

●面向车载视觉的迁移学习和领域适应

●基于虚拟图像的车载视觉系统评价

●将虚拟图像用于智能车辆的其他研究

截稿日期：2018年1月31日（以官网时间为准）

组织委员会：

王坤峰副研究员

苟超助理研究员

David Vázquez博士

Hui Yu博士

Workshop网站地址:

The 2018 IEEE Intelligent Vehicles Symposium (IV'18)

Changshu,Jiangsu,China

June 26 - June 30, 2018

Call for Papers

IV’18 Workshop onParallel Vision in Intelligent Vehicles (Code: 319nv)

Content and List of Topics

Recent advances in computer vision (including object detection, segmentation, tracking, and scene understanding) have promoted the development of intelligent vehicles significantly. However, vision-based methods for intelligent vehicles research require large amounts of annotated images for training, testing, and explaining the computer vision models effectively, while the collection of large-scale, diversified labelled data from the real world is both expensive and error-prone. To tackle these issues, some researchers at Chinese Academy of Sciences proposed parallel vision, which is a virtual-real interactive approach to intelligent visual computing and comprises artificial scenes, computational experiments, and parallel execution. By synthesizing and exploiting virtual images to build more reliable computer vision systems, parallel vision is attracting increasingly more attention in the community of computer vision and intelligent vehicles.

This workshop aims at getting researchers of parallel vision together and stimulating discussion on its utility in intelligent vehicles.The list of possible topics includes, but is not limited to:

●Computer graphics and virtual reality for traffic scene simulation

●Driving simulator to generate photorealistic virtual images

●Generative models related to virtual traffic images(generative adversarial networks, variational autoencoders, etc.)

●Neural networks that learn from virtual images

●Intelligent visual computing with virtual images

●Virtual and real world adaptation for vehicular vision

●Evaluation of vehicular vision systems using virtual images

●Applications of virtual images to intelligent vehicles

Submission Guidelines

Important Dates

Submission Deadline:January 31, 2018

Notification of paper acceptance:April 15, 2018

Final paper submission:April 25, 2018

Organizers