Experiment Question

[eds89]

Hello.
I've read the paper and I've got a few questions about the Imitation Learning experiment.

I'm not a deep learning expert, so some questions may sound silly.

1. When executing the navigation tasks, after training, do the trained system receives new images from the camera installed on the car? Do you still have the images recorded during the execution of the expriment?

2. What camera mode was used by the forward-facing camera? Also, what there any kind of post-processing applied to the camera (i.e. ground truth, semantic segmentation, or just scene final)?

[felipecode]

Hey @eds89 , no question is really silly :).

1. Yes, during evaluation you need to have a CARLA running. Then you will receive new images and control de car. The recorded images are only for training the neural network.

2. The forward facing camera was just RGB post-processed (scene-final), no depth or semantic segmentation was used.

[eds89]

A follow-up question: 1) How can I change the experiment to have a different camera setting while moving to the New Town (after the net is trained in Town01)? Since the new camera will inevitably receive data, for which the net had not been specifically trained for, I can already predict there will be changes in the obtained results. I.e. more infractions or even inability to drive the car at all. Iwant to know how much the new results differ from the original ones.

…

On 5 February 2018 at 08:40, felipecode ***@***.***> wrote: Hey @eds89 <https://github.com/eds89> , no question is really silly :). 1. Yes, during evaluation you need to have a CARLA running. Then you will receive new images and control de car. The recorded images are only for training the neural network. 2. The forward facing camera was just RGB post-processed (scene-final), no depth or semantic segmentation was used. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#6 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AQUNuS9NQRAQZQXf9a0_KR4L-XfqmDF2ks5tRr6agaJpZM4R3VCH> .