Autonomous Driving Analysis, sumarry And Conclusion - Versionsgeschichte

Evrard.leuteu-feukeu@stud.hshl.de am 12. Juni 2025 um 21:36 Uhr

2025-06-12T21:36:47Z

← Nächstältere Version		Version vom 12. Juni 2025, 21:36 Uhr
Zeile 6:		Zeile 6:
	* The project switched to classification using discrete steering values. with the following parameters as seen on the figure below:		* The project switched to classification using discrete steering values. with the following parameters as seen on the figure below:

	[[Datei:Cleaned CNN Classify.png\|~~mini~~\|Fig. 38: Classification Training]]		[[Datei:Cleaned CNN Classify.png\|\|Fig. 38: Classification Training]]
	[[Datei:regression.png\|~~mini~~\|Fig. 39: Classification Result]]		[[Datei:regression.png\|\|Fig. 39: Classification Result]]

	More details of further work on autonomous driving can be found here[https://wiki.hshl.de/wiki/index.php/JetRacer:_Autonomous_Driving,_Obstacle_Detection_and_Avoidance_using_AI_with_MATLAB].		More details of further work on autonomous driving can be found here[https://wiki.hshl.de/wiki/index.php/JetRacer:_Autonomous_Driving,_Obstacle_Detection_and_Avoidance_using_AI_with_MATLAB].

Evrard.leuteu-feukeu@stud.hshl.de am 12. Juni 2025 um 20:59 Uhr

2025-06-12T20:59:48Z

Evrard.leuteu-feukeu@stud.hshl.de am 12. Juni 2025 um 19:39 Uhr

2025-06-12T19:39:40Z

← Nächstältere Version		Version vom 12. Juni 2025, 19:39 Uhr
Zeile 6:		Zeile 6:
	* The project switched to classification using discrete steering values. with the following parameters as seen on the figure below:		* The project switched to classification using discrete steering values. with the following parameters as seen on the figure below:

	[[Datei:Cleaned CNN Classify.png\|mini\|Fig. 13: ~~Camera setup for image recording during driving~~]]		[[Datei:Cleaned CNN Classify.png\|mini\|Fig. 38: Classification Training]]
	[[Datei:regression.png\|mini\|Fig. 5: ~~Camera setup for image without calibration~~]]		[[Datei:regression.png\|mini\|Fig. 39: Classification Result]]

	More details of further work on Autonomous driving can be found here[https://wiki.hshl.de/wiki/index.php/JetRacer:_Autonomous_Driving,_Obstacle_Detection_and_Avoidance_using_AI_with_MATLAB].		More details of further work on Autonomous driving can be found here[https://wiki.hshl.de/wiki/index.php/JetRacer:_Autonomous_Driving,_Obstacle_Detection_and_Avoidance_using_AI_with_MATLAB].

Evrard.leuteu-feukeu@stud.hshl.de am 12. Juni 2025 um 19:07 Uhr

2025-06-12T19:07:30Z

← Nächstältere Version		Version vom 12. Juni 2025, 19:07 Uhr
Zeile 6:		Zeile 6:
	* The project switched to classification using discrete steering values. with the following parameters as seen on the figure below:		* The project switched to classification using discrete steering values. with the following parameters as seen on the figure below:

	[[Datei:Cleaned CNN Classify.png\|mini~~\|frame~~\|Fig. 13: Camera setup for image recording during driving]]		[[Datei:Cleaned CNN Classify.png\|mini\|Fig. 13: Camera setup for image recording during driving]]
	[[Datei:regression.png\|mini]]		[[Datei:regression.png\|mini\|Fig. 5: Camera setup for image without calibration]]

	More details of further work on Autonomous driving can be found here[https://wiki.hshl.de/wiki/index.php/JetRacer:_Autonomous_Driving,_Obstacle_Detection_and_Avoidance_using_AI_with_MATLAB].		More details of further work on Autonomous driving can be found here[https://wiki.hshl.de/wiki/index.php/JetRacer:_Autonomous_Driving,_Obstacle_Detection_and_Avoidance_using_AI_with_MATLAB].

Evrard.leuteu-feukeu@stud.hshl.de am 12. Juni 2025 um 19:04 Uhr

2025-06-12T19:04:35Z

← Nächstältere Version		Version vom 12. Juni 2025, 19:04 Uhr
Zeile 24:		Zeile 24:
	* However, simultaneous recording, steering, and saving caused performance bottlenecks due to the Jetson Nano’s limited resources.		* However, simultaneous recording, steering, and saving caused performance bottlenecks due to the Jetson Nano’s limited resources.

	Note: ~~that i~~ have provided alternatives and solutions to the limitations above and could not ~~impliment~~ them myself due to time ~~constraint~~.		Note: I have provided alternatives and solutions to the limitations above and could not implement them myself due to time constraints.


Zeile 33:		Zeile 33:

	This experiment aimed to use the Jetson Nano and Matlab to complete various tasks from "Familiarization with the existing framework" to "Autonomous Driving". Although the final result of the car driving several labs in the right lane could not be attained, there is plenty of room for future improvements as far as:		This experiment aimed to use the Jetson Nano and Matlab to complete various tasks from "Familiarization with the existing framework" to "Autonomous Driving". Although the final result of the car driving several labs in the right lane could not be attained, there is plenty of room for future improvements as far as:
	* Training model method with (more ~~datas~~, different training ~~algorithm~~ like reinforcement learning etc)		* Training model method with (more data, different training algorithms like reinforcement learning etc)
	* Expanding the dataset with more diverse environmental conditions.		* Expanding the dataset with more diverse environmental conditions.
	* Flash working Model directly on jetson nano using ~~gpu coder~~		* Flash working Model directly on jetson nano using GPU Coder.
	* Use accurate gamepad.		* Use accurate gamepad.
	* Improving automated labeling and refining the PD controller parameters for faster driving without loss of robustness.		* Improving automated labeling and refining the PD controller parameters for faster driving without loss of robustness.

Evrard.leuteu-feukeu@stud.hshl.de am 12. Juni 2025 um 19:02 Uhr

2025-06-12T19:02:23Z

← Nächstältere Version		Version vom 12. Juni 2025, 19:02 Uhr
Zeile 19:		Zeile 19:
	=== Limitations ===		=== Limitations ===
	* Hardware constraints limited the maximum processing speed(without GPU Coder).		* Hardware constraints limited the maximum processing speed(without GPU Coder).
			* Camera did not allow 50 fps and recording dimension was limited to 256 x 256.
	* Initial data quality was poor due to randomized manual steering input.		* Initial data quality was poor due to randomized manual steering input.
	* Better results were obtained after switching to discrete, consistent control signals.		* Better results were obtained after switching to discrete, consistent control signals.
	* However, simultaneous recording, steering, and saving caused performance bottlenecks due to the Jetson Nano’s limited resources.		* However, simultaneous recording, steering, and saving caused performance bottlenecks due to the Jetson Nano’s limited resources.

			Note: that i have provided alternatives and solutions to the limitations above and could not impliment them myself due to time constraint.

Evrard.leuteu-feukeu@stud.hshl.de am 12. Juni 2025 um 18:56 Uhr

2025-06-12T18:56:13Z

@@ Zeile 1: / Zeile 1: @@
 == Autonomous Driving Neural Network Training ==
 == Experiment Analysis ==
@@ Zeile 13: / Zeile 25: @@
 == Summary and Outlook ==

Evrard.leuteu-feukeu@stud.hshl.de am 12. Juni 2025 um 18:55 Uhr

2025-06-12T18:55:16Z

← Nächstältere Version		Version vom 12. Juni 2025, 18:55 Uhr
Zeile 22:		Zeile 22:
	[[Datei:regression.png\|mini]]		[[Datei:regression.png\|mini]]

	More details of further work on Autonomous driving can be found here[].		More details of further work on Autonomous driving can be found here[https://wiki.hshl.de/wiki/index.php/JetRacer:_Autonomous_Driving,_Obstacle_Detection_and_Avoidance_using_AI_with_MATLAB].

	== Summary and Outlook ==		== Summary and Outlook ==

Evrard.leuteu-feukeu@stud.hshl.de am 12. Juni 2025 um 18:53 Uhr

2025-06-12T18:53:29Z

← Nächstältere Version		Version vom 12. Juni 2025, 18:53 Uhr
Zeile 1:		Zeile 1:
	== Autonomous Driving Neural Network Training ==		== Autonomous Driving Neural Network Training ==
	== Experiment Analysis ==		== Experiment Analysis ==
	* Models trained on discrete steering values showed significantly better results than regression-based approaches.
	* Grayscale recordings increased model stability.
	* Calibrated images helped reduce the steering error caused by lens distortion.


Zeile 17:		Zeile 14:

	=== Training Results ===		=== Training Results ===
	* The project switched to classification using discrete steering values. with the following parameters:		* Models trained on discrete steering values showed significantly better results than regression-based approaches.
			* Grayscale recordings increased model stability.
			* Calibrated images helped reduce the steering error caused by lens distortion.
			* The project switched to classification using discrete steering values. with the following parameters as seen on the figure below:

	[[Datei:Cleaned CNN Classify.png\|mini\|frame\|Fig. 13: Camera setup for image recording during driving]]		[[Datei:Cleaned CNN Classify.png\|mini\|frame\|Fig. 13: Camera setup for image recording during driving]]
	[[Datei:regression.png\|mini]]		[[Datei:regression.png\|mini]]
Zeile 25:		Zeile 26:
	== Summary and Outlook ==		== Summary and Outlook ==

	This ~~project demonstrated that it is possible~~ to ~~optimize autonomous navigation on~~ the ~~JetRacer using a combination of classical control algorithms~~ and ~~neural networks~~. ~~Despite hardware limitations~~, ~~stable autonomous driving behavior was achieved.~~		This experiment aimed to use the Jetson Nano and Matlab to complete various tasks from "Familiarization with the existing framework" to "Autonomous Driving". Although the final result of the car driving several labs in the right lane could not be attained, there is plenty of room for future improvements as far as:
			* Training model method with (more datas, different training algorithm like reinforcement learning etc)
	~~Future~~ improvements ~~could include~~:
	* Expanding the dataset with more diverse environmental conditions.		* Expanding the dataset with more diverse environmental conditions.
	* Flash working Model directly on jetson nano using gpu coder		* Flash working Model directly on jetson nano using gpu coder
	* Use accurate gamepad ~~and use more complex training method~~.		* Use accurate gamepad.
	* Improving automated labeling and refining the PD controller parameters for faster driving without loss of robustness.		* Improving automated labeling and refining the PD controller parameters for faster driving without loss of robustness.

Evrard.leuteu-feukeu@stud.hshl.de am 12. Juni 2025 um 18:40 Uhr

2025-06-12T18:40:41Z

← Nächstältere Version		Version vom 12. Juni 2025, 18:40 Uhr
Zeile 17:		Zeile 17:

	=== Training Results ===		=== Training Results ===
	* Pretrained methods were tried but also showed poor performance. with the following parameters:
	~~[[Datei:Pretrained 50 epoch.png\|mini]]~~
	~~[[Datei:Pretrained 50 epoch 2.png\|mini]]~~

	* Regression methods were tried but also showed poor performance. with the following parameters:
	~~[[Datei:regression_50_epoch2.png\|mini\|Fig. 9: regression_50_epoch2]]~~
	~~[[Datei:regression_50_epoch.png\|mini\|frame\|Fig. 10: regression_50_epoch]]~~

	* The project switched to classification using discrete steering values. with the following parameters:		* The project switched to classification using discrete steering values. with the following parameters:
	[[Datei:Cleaned CNN Classify.png\|mini\|frame\|Fig. 13: Camera setup for image recording during driving]]		[[Datei:Cleaned CNN Classify.png\|mini\|frame\|Fig. 13: Camera setup for image recording during driving]]

← Nächstältere Version		Version vom 12. Juni 2025, 20:59 Uhr
Zeile 9:		Zeile 9:
	[[Datei:regression.png\|mini\|Fig. 39: Classification Result]]		[[Datei:regression.png\|mini\|Fig. 39: Classification Result]]

	More details of further work on ~~Autonomous~~ driving can be found here[https://wiki.hshl.de/wiki/index.php/JetRacer:_Autonomous_Driving,_Obstacle_Detection_and_Avoidance_using_AI_with_MATLAB].		More details of further work on autonomous driving can be found here[https://wiki.hshl.de/wiki/index.php/JetRacer:_Autonomous_Driving,_Obstacle_Detection_and_Avoidance_using_AI_with_MATLAB].


Zeile 18:		Zeile 18:

	=== Limitations ===		=== Limitations ===
	* Hardware constraints limited the maximum processing speed(without GPU Coder).		* Hardware constraints limited the maximum processing speed (without GPU Coder).
	* ~~Camera~~ did not allow 50 fps and recording dimension was limited to 256 x 256.		* The camera did not allow 50 fps, and the recording dimension was limited to 256 x 256.
	* Initial data quality was poor due to randomized manual steering input.		* Initial data quality was poor due to randomized manual steering input.
	* Better results were obtained after switching to discrete, consistent control signals.		* Better results were obtained after switching to discrete, consistent control signals.
	* However, simultaneous recording, steering, and saving caused performance bottlenecks due to the Jetson Nano’s limited resources.		* However, simultaneous recording, steering, and saving caused performance bottlenecks due to the Jetson Nano’s limited resources.

	Note: I have provided alternatives and solutions to the limitations above and could not implement them myself due to time constraints.		Note: I have provided alternatives and solutions to the limitations above and could not implement them myself due to time constraints.


Zeile 32:		Zeile 32:
	== Summary and Outlook ==		== Summary and Outlook ==

	This experiment aimed to use the Jetson Nano and Matlab to complete various tasks from "Familiarization with the existing framework" to "Autonomous Driving". Although the final result of the car driving several ~~labs~~ in the right lane could not be attained, there is plenty of room for future improvements as far as:		This experiment aimed to use the Jetson Nano and Matlab to complete various tasks, from "Familiarization with the existing framework" to "Autonomous Driving." Although the final result of the car driving several laps in the right lane could not be attained, there is plenty of room for future improvements as far as:
	* Training model method with (more data, different training algorithms like reinforcement learning etc)		* Training model method with (more data, different training algorithms like reinforcement learning, etc.)
	* Expanding the dataset with more diverse environmental conditions.		* Expanding the dataset with more diverse environmental conditions.
	* Flash working ~~Model~~ directly on ~~jetson nano~~ using GPU Coder.		* Flash working model directly on Jetson Nano using GPU Coder.
	* Use accurate gamepad.		* Use an accurate gamepad.
	* Improving automated labeling and refining the PD controller parameters for faster driving without loss of robustness.		* Improving automated labeling and refining the PD controller parameters for faster driving without loss of robustness.