Geodesic
Enhanced Image Preprocessing Method for an Autonomous Vehicle Agent System
Computer Science and Information Systems, Tome 18 (2021) no. 2.

Voir la notice de l'article provenant de la source Computer Science and Information Systems website

Excessive training time is a major issue face when training autonomous vehicle agents with neural networks by using images as input. This paper proposes a deep time-economical Q network (DQN) input image preprocessing method to train an autonomous vehicle agent in a virtual environment. The environmental information is extracted from the virtual environment. A top-view image of the entire environment is then redrawn according to the environmental information. During training of the DQN model, the top-view image is cropped to place the vehicle agent at the center of the cropped image. The current frame top-view image is combined with the images from the previous two training iterations. The DQN model use this combined image as input. The experimental results indicate higher performance and shorter training time for the DQN model trained with the preprocessed images compared with that trained without preprocessing.
Keywords: Image preprocessing, Reinforcement learning, Deep Q learning 
@article{CSIS_2021_18_2_a5,
     author = {Kaisi Huang and Mingyun Wen and Jisun Park and Yunsick Sung and Jong Hyuk Park and Kyungeun Cho},
     title = {Enhanced {Image} {Preprocessing} {Method} for an {Autonomous} {Vehicle} {Agent} {System}},
     journal = {Computer Science and Information Systems},
     publisher = {mathdoc},
     volume = {18},
     number = {2},
     year = {2021},
     url = {http://geodesic.mathdoc.fr/item/CSIS_2021_18_2_a5/}
}
TY  - JOUR
AU  - Kaisi Huang
AU  - Mingyun Wen
AU  - Jisun Park
AU  - Yunsick Sung
AU  - Jong Hyuk Park
AU  - Kyungeun Cho
TI  - Enhanced Image Preprocessing Method for an Autonomous Vehicle Agent System
JO  - Computer Science and Information Systems
PY  - 2021
VL  - 18
IS  - 2
PB  - mathdoc
UR  - http://geodesic.mathdoc.fr/item/CSIS_2021_18_2_a5/
ID  - CSIS_2021_18_2_a5
ER  - 
%0 Journal Article
%A Kaisi Huang
%A Mingyun Wen
%A Jisun Park
%A Yunsick Sung
%A Jong Hyuk Park
%A Kyungeun Cho
%T Enhanced Image Preprocessing Method for an Autonomous Vehicle Agent System
%J Computer Science and Information Systems
%D 2021
%V 18
%N 2
%I mathdoc
%U http://geodesic.mathdoc.fr/item/CSIS_2021_18_2_a5/
%F CSIS_2021_18_2_a5
Kaisi Huang; Mingyun Wen; Jisun Park; Yunsick Sung; Jong Hyuk Park; Kyungeun Cho. Enhanced Image Preprocessing Method for an Autonomous Vehicle Agent System. Computer Science and Information Systems, Tome 18 (2021) no. 2. http://geodesic.mathdoc.fr/item/CSIS_2021_18_2_a5/