A Survey on Theories and Applications for Self-Driving Cars Based on Deep Learning Methods

: Self-driving cars are a hot research topic in science and technology, which has a great inﬂuence on social and economic development. Deep learning is one of the current key areas in the ﬁeld of artiﬁcial intelligence research. It has been widely applied in image processing, natural language understanding, and so on. In recent years, more and more deep learning-based solutions have been presented in the ﬁeld of self-driving cars and have achieved outstanding results. This paper presents a review of recent research on theories and applications of deep learning for self-driving cars. This survey provides a detailed explanation of the developments of self-driving cars and summarizes the applications of deep learning methods in the ﬁeld of self-driving cars. Then the main problems in self-driving cars and their solutions based on deep learning methods are analyzed, such as obstacle detection, scene recognition, lane detection, navigation and path planning. In addition, the details of some representative approaches for self-driving cars using deep learning methods are summarized. Finally, the future challenges in the applications of deep learning for self-driving cars are given out.


Introduction
Self-driving cars are a much discussed topic today.It is also known as autonomous vehicles (AVs) or driverless cars, represents a transformative technology poised to revolutionize the way we travel.These vehicles, equipped with advanced sensors, artificial intelligence (AI), and communication systems, have the potential to enhance road safety, improve traffic flow, and increase mobility for individuals worldwide.
Self-driving cars have attracted more and more attention due to their significant economic impact.However there are a lots of challenges in self-driving cars , Example; safety problem is the key technology that must be solved efficiently in self-driving cars, otherwise, it is impossible to allow self-driving cars on the road.This paper provide a survey on theories and applications of deep learning for self-driving cars.Other relevant surveys in the field of deep learning and self-driving cars can be used as a supplement to this paper.

Related Information
The concept of autonomous vehicles traces back to the early 20th century, with pioneers like Leonardo da Vinci envisioning self-propelled vehicles.However, significant advancements in the field began in the latter half of the 20th century.In 1987, the U.S. Congress established the Defense Advanced Research Projects Agency (DARPA) to develop autonomous vehicle technology for military purposes.This led to the creation of the DARPA Grand Challenge in the early 2000s, spurring innovation in the private sector.
The levels of autonomy in cars, as defined by the Society of Automotive Engineers (SAE) International, are commonly referred to as the SAE J3016 levels.These levels provide a framework for categorizing the degree of automation in vehicles, ranging from no automation to full automation.Here's an overview of the SAE J3016 levels: Level 0, represents a vehicle with no autonomy.Level 1, has basic driving assists like adaptive cruise and emergency braking.Level 2, consists of partial autonomy while requiring the driver to monitor the system and perform certain tasks.Level 3, Under certain conditions the system has full autonomy , but a human operator is still required to take control if necessary.Level 4, The vehicles at Level 4 is still a semiautonomous system with more automation than Level 3. Level 5, The vehicles are fully autonomous in all circumstances.
Remark 1 (About Level 5 -Fully autonomous).Level 5, vehicles do not require steering wheels, pedals, or other manual controls, as there is no provision for human driving.These vehicles offer complete autonomy and are designed to operate safely and efficiently without any input from human occupants.These levels provide a standardized framework for understanding the capabilities and limitations of autonomous vehicles, helping to guide the development, testing, and deployment of self-driving technology.Self-driving cars offer a range of potential benefits, including: 1. Safety improvements; 2. Traffic efficiency and congestion reduction.; 3. Environmental impact; 4. Accessibility.

Remark 2 (About Tesla Self-driving cars). "Auto
Self-driving cars rely on a combination of sensors, including cameras, LiDAR (Light Detection and Ranging), radar, and GPS, to perceive their surroundings.These sensors provide real-time data about nearby objects, road conditions, and traffic patterns, enabling the vehicle to make informed decisions.AI algorithms process this data to navigate the vehicle, anticipate potential hazards, and execute driving maneuvers safely.
The overall technical framework of self-driving cars with a level3 or higher autonomy system is divided into four parts ; First is the driving environment perception system, the autonomous decision system, the control execution system and the monitor system.The architecture of it is shown in Figure 2.
Remark 3; According to the autonomous driving levels, the car from Label 0 to Label 2 mainly requires the driver to monitor the environment .Advanced Driver Assistant System (ADAS) are intelligent systems that reside inside vehicles classified from Level 0 to Level 2 and assist the driver in the driving process.

Safety
Ensuring the safe operation of self-driving cars to for the safety of passengers and other road users..

Privacy
Addressing concerns around data collection and use in autonomous vehicles.

Liability
Defining clear legal framework for liability in autonomous vehicle accidents.

Figure 1 .
Figure 1.Levels of autonomous driving according to SAE J3016 -pilot" technology has made major breakthroughs in recent years.Although the tesla's autopilot technology is only regarded as Level2 stage by the national highway traffic safety administration (NHTSA), as one of the most successful companies in autopilot system Lavanya Chauhan, IJSRM Volume 12 Issue 04 April 2024 EC-2024-1161application by far, Tesla shows us that the car has basically realized automatic driving under certain conditions (see Figure2b).

Figure 3 .
Figure 3 .The overall technical framework of self-driving cars with a level3 or higher autonomy system