Evolution of an artificial neural network based autonomous land vehicle controller

By Shumeet Baluja. 1996

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

A Survey of Autonomous Driving: Common Practices and Emerging Technologies (2019, methodology): “ALVINN was trained with neuroevolution and outperformed the direct supervised learning version [62].”
Genetic Representation and Evolvability of Modular Neural Controllers (2010, background): “However, in many robotic tasks where multiple interdependent control problems have to be solved simultaneously, the performance of conventional neuroevolution techniques declines.”
Evolution of homing navigation in a real mobile robot (1996, methodology): “Baluja shows that genetic algorithms can provide strategies to control an autonomous land vehicle (ALVINN) that are comparable to those found by a supervised learning algorithm [65].”
Neurovisual Control in the Quake II Environment (2012, methodology): “Baluja modified the experiment by training the weights of the NN using an evolutionary computational model [16].”
Color-based road detection in urban traffic scenes (2004, methodology): “Some methods exist that are based supervised learning, which always take a lot of time in the training and testing steps [ 27 ], [28] and make little use of the existing features of the road in urban and campus environments.”
Video-based lane estimation and tracking for driver assistance: survey, system, and evaluation (2006, methodology): “ALVINN (1996) [19], [36] A Flat road model for generating training data Image intensity Neural Network None Road tests, various error measure associated with neural networks Neural network makes it difficult to decouple control from detection, requires lots of training RALPH (1996) [25] A B Constant”
Finding multiple lanes in urban road networks with vision and lidar (2009, unknown): “The resulting road boundary detections are used as evidence and incorporated into an evidence grid, a process discussed in the following section.”