Sora-Based Parallel Vision for Smart Sensing of Intelligent Vehicles: From Foundation Models to Foundation Intelligence

IF 14 1区工程技术 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

IEEE Transactions on Intelligent Vehicles Pub Date : 2024-02-01 DOI:10.1109/TIV.2024.3376575

Hongkai Yu;Xinyu Liu;Yonglin Tian;Yutong Wang;Chao Gou;Fei-Yue Wang

引用次数: 0

Abstract

There are a large number of functional sensors installed on the modern intelligent vehicles. Many Artificial Intelligence based foundation models have been proposed for smart sensing to recognize the known object classes in the new but similar scenarios. However, it is still challenging for the foundation models of smart sensing to detect all the object classes in both seen and unseen scenarios. This letter aims at pushing the boundary of smart sensing research for intelligent vehicles. We first summarize the current widely-used foundation models and the foundation intelligence needed for smart sensing of intelligent vehicles. We then explain Sora-based Parallel Vision to boost the foundation models of smart sensing from basic intelligence (1.0) to enhanced intelligence (2.0) and final generalized intelligence (3.0). Several representative case studies are discussed to show the potential usages of Sora-based Parallel Vision, followed by its future research direction.

查看原文本刊更多论文

基于 Sora 的并行视觉，用于智能车辆的智能传感：从基础模型到基础智能

现代智能车辆上安装了大量功能传感器。人们提出了许多基于人工智能的智能传感基础模型，以便在新的但相似的场景中识别已知的物体类别。然而，智能传感的基础模型要在看到和看不到的场景中检测到所有物体类别，仍然具有挑战性。这封信旨在推动智能车辆的智能传感研究。我们首先总结了目前广泛使用的基础模型以及智能车辆智能感知所需的基础智能。然后，我们解释了基于 Sora 的并行视觉如何将智能感知的基础模型从基本智能（1.0）提升到增强智能（2.0）和最终的通用智能（3.0）。我们还讨论了几个具有代表性的案例研究，以展示基于 Sora 的并行视觉的潜在用途，以及其未来的研究方向。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

IEEE Transactions on Intelligent Vehicles Mathematics-Control and Optimization

CiteScore

12.10

自引率

13.40%

发文量

177

期刊介绍： The IEEE Transactions on Intelligent Vehicles (T-IV) is a premier platform for publishing peer-reviewed articles that present innovative research concepts, application results, significant theoretical findings, and application case studies in the field of intelligent vehicles. With a particular emphasis on automated vehicles within roadway environments, T-IV aims to raise awareness of pressing research and application challenges. Our focus is on providing critical information to the intelligent vehicle community, serving as a dissemination vehicle for IEEE ITS Society members and others interested in learning about the state-of-the-art developments and progress in research and applications related to intelligent vehicles. Join us in advancing knowledge and innovation in this dynamic field.