Open howthep opened 1 month ago
Hello! The evaluation metrics method is provided in part by the autonomous exploration framework provided by CMU (see Code repository: https://github.com/HongbiaoZ/autonomous_exploration_development_environment and their website: https://www.cmu-exploration.com), which provides the complete code. In fact, we work within this framework. Please feel free to contact us if you have any questions.
Hi, thanks for your response!
I noticed that the framework includes a visualization package, and it publishes the explored volume. However, the paper mentions that the evaluation metric is based on the explored area.
Does this mean that to obtain the area, I should divide the volume by a certain height or use some other method? Could you provide more details or clarification on how the area is calculated from the volume?
Thanks again for your help!
Hi! If you want to obtain the area of the explored environment, you can obtain it by multiplying the number of known grids (both free and occupied) by the area of a single grid. This may require you to write a script implementation. In fact, the size of each grid can be determined by the grid resolution (say, 0.5m).
Or you can just use the volume you've already explored instead of the area, which essentially doesn't matter, because exploration is also the process of exploring a 3d environment.
Sorry for my unfamiliarity with the framework, but I’m not sure which topic publishes the grid. I did find a 2D grid map—could this be what I need for calculating the area?
I’d appreciate any guidance on this!
Hi, thank you for your work!
I’m particularly interested in the evaluation metrics mentioned in the paper. I was wondering if there is any available code or scripts to reproduce the evaluation metrics as described.
If the code is not available, could you provide guidance or any references on how to implement these metrics?
Thanks in advance for your help!