Surround-View Cameras based Holistic Visual Perception for Automated Driving

TL;DR

This paper proposes a holistic visual perception approach using surround-view cameras for automated driving, focusing on near-field perception with high performance and low computational complexity, leveraging multi-task learning to optimize efficiency.

Contribution

It introduces near-field perception algorithms with multi-task learning to improve real-time performance in automated driving systems.

Findings

Developed high-performance, low-complexity perception algorithms.

Applied multi-task learning to share convolutional layers across tasks.

Achieved efficient near-field perception suitable for real-time applications.

Abstract

The formation of eyes led to the big bang of evolution. The dynamics changed from a primitive organism waiting for the food to come into contact for eating food being sought after by visual sensors. The human eye is one of the most sophisticated developments of evolution, but it still has defects. Humans have evolved a biological perception algorithm capable of driving cars, operating machinery, piloting aircraft, and navigating ships over millions of years. Automating these capabilities for computers is critical for various applications, including self-driving cars, augmented reality, and architectural surveying. Near-field visual perception in the context of self-driving cars can perceive the environment in a range of $0-10$ meters and 360{\deg} coverage around the vehicle. It is a critical decision-making component in the development of safer automated driving. Recent advances in computer vision and deep learning, in conjunction with high-quality sensors such as cameras and LiDARs, have fueled mature visual perception solutions. Until now, far-field perception has been the primary focus. Another significant issue is the limited processing power available for developing real-time applications. Because of this bottleneck, there is frequently a trade-off between performance and run-time efficiency. We concentrate on the following issues in order to address them: 1) Developing near-field perception algorithms with high performance and low computational complexity for various visual perception tasks such as geometric and semantic tasks using convolutional neural networks. 2) Using Multi-Task Learning to overcome computational bottlenecks by sharing initial convolutional layers between tasks and developing optimization strategies that balance tasks.