sisl/2016_itsc_scenegen

Model details, parameters, and associated code for the scene generation models associated with "Factor Graph Scene Distributions for Automotive Safety Analysis" - ITSC 2016

2016_itsc_scenegen's Issues

Final Touches before Final Submission

Reviewer Comments

Reviewer 8

• It would be interesting to see the performance with more factor features and more complicated
  traffic networks.

mentioned this in the conclusion

Reviewer 10

• In the first page left column the beginning of the second paragraph in the introduction section the authors said "Estimated metrics depend strongly on the distribution of scenes represented by the model" which is a strong statement and they did not give any reference.
• The colors used in figure 2 and figure 4 look similar and the reader may not be able to differentiate between them.

Changing factor circles to include custom Tikz contents so they can be distinguished in fig 2. Changed around the colors and line styles in fig 4.

• The authors need to indicates what is the advantage of a simulator using the proposed scene generation approach and the state of practice simulators? Since both the proposed scene generation method and the state of practice simulators need burn in.

Advantage would have been that you have a statistically valid distribution over scenes - you don't get that through sim unless you know what sorts of vehicles to spawn where / what behaviors to assign to them. I now mention this in the intro.

• The authors need to indicate how salable the proposed scene generator and is it applicable for large networks?

Mentioned in conclusion

Misc

• Cut down to 6 pages
• Even out references
• Final spellcheck

Author Checkout

• Mykel
• Philipp
• Tim