Training on GOOSE Dataset

[rapzag]

Thanks for this source code! We recently open sourced first parts of the GOOSE dataset and I trained GANav on the same.

The qualitative results look quite promising but the numbers appear quite low compared to your results on RUGD and Rellis3D. Of course this is difficult to compare as no SOTA mIOU is established yet.

+-------+-------+-------+
|  aAcc |  mIoU |  mAcc |
+-------+-------+-------+
| 57.27 | 36.64 | 51.84 |
+-------+-------+-------+
+---------------------+-------+-------+
|        Class        |  IoU  |  Acc  |
+---------------------+-------+-------+
| background/obstacle | 71.94 | 73.34 |
|        stable       |  34.7 | 85.21 |
|       granular      | 15.96 | 23.93 |
|    poor foothold    | 21.94 | 32.33 |
|   high resistance   | 17.66 |  21.3 |
|         none        | 57.67 | 74.96 |
+---------------------+-------+-------+

I took the 6-class approach, with the following categorization and otherwise default parameters:

# 0 background: sky
# 1 Stable: bikeway, pedesstrian_crossing, road_marking, sidewalk, asphalt,
# 2 Granular: cobble, leaves, moss, gravel, soil
# 3 Poor foothold: snow, low_grass,
# 4 High resistance: high_grass, bush, debris, crops, water, tree_root
# 5 Obstacle: everything else

Do you have any tips for fine-tuning the training? If of interest, I could also add a PR with the config for GOOSE.

GANav-goose.mp4

[rayguan97]

Hi,

Welcome to submit a PR to incorporate GOOSE dataset into the config and preparation steps in readme!

Since I have not worked on this dataset, I do not have any experience or general tips on this dataset. You are also welcome to send me an email so we can discuss it further, once I have more details about this dataset.

Best,
Tianrui

[rapzag]

There you go, I added some basic configuration. If you have time to try the training configuration, I would greatly appreciate some feedback on good parameters / mappings.

Best,
Raphael

[rayguan97]

Thank you for the pull request. It has been integrated into the repo. :)