Update README.md
Browse files
README.md
CHANGED
|
@@ -2,9 +2,25 @@
|
|
| 2 |
tags:
|
| 3 |
- model_hub_mixin
|
| 4 |
- pytorch_model_hub_mixin
|
|
|
|
| 5 |
---
|
| 6 |
|
| 7 |
-
|
| 8 |
-
|
| 9 |
-
|
| 10 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
tags:
|
| 3 |
- model_hub_mixin
|
| 4 |
- pytorch_model_hub_mixin
|
| 5 |
+
pipeline_tag: robotics
|
| 6 |
---
|
| 7 |
|
| 8 |
+
PointGoal navigation policy trained with the FrodoBots8K dataset.
|
| 9 |
+
|
| 10 |
+
Presented in the paper [Data scaling for navigation in unknown environments](https://arxiv.org/abs/2601.09444)
|
| 11 |
+
|
| 12 |
+
[Project page](https://lasuomela.github.io/navigation_scaling/) | [Code](https://github.com/lasuomela/NavigationScaling)
|
| 13 |
+
|
| 14 |
+
## Details
|
| 15 |
+
|
| 16 |
+
Architecture:
|
| 17 |
+
|
| 18 |
+
MLP-BC with Theia encoder + MLP action head
|
| 19 |
+
|
| 20 |
+
Dataset:
|
| 21 |
+
|
| 22 |
+
512H total, 8H from 64 distinct geographical locations
|
| 23 |
+
|
| 24 |
+
## Usage
|
| 25 |
+
|
| 26 |
+
See the Github repo
|