Update README.md
Browse files
README.md
CHANGED
|
@@ -1,11 +1,6 @@
|
|
| 1 |
---
|
|
|
|
| 2 |
license: gpl-3.0
|
| 3 |
-
tags:
|
| 4 |
-
- robotics
|
| 5 |
-
- lidar
|
| 6 |
-
- legged-locomotion
|
| 7 |
-
- imu
|
| 8 |
-
- unitree-go2
|
| 9 |
---
|
| 10 |
|
| 11 |
# JEPLO Dataset
|
|
@@ -18,6 +13,8 @@ This dataset accompanies the paper **JEPLO: Joint-Embedding Predictive Learning
|
|
| 18 |
|
| 19 |
**Code:** [ASIG-X/JEPLO](https://github.com/ASIG-X/JEPLO)
|
| 20 |
|
|
|
|
|
|
|
| 21 |
## Dataset Overview
|
| 22 |
|
| 23 |
Publicly released to support the legged robotics community, the dataset comprises **eight sequences** recorded during our experiments: **five indoor sequences** and **three mixed indoor–outdoor sequences**.
|
|
@@ -29,6 +26,7 @@ The sequences were recorded with a Unitree Go2 quadruped equipped with a Livox M
|
|
| 29 |
- Joint encoder measurements.
|
| 30 |
- 6-DoF ground-truth trajectories.
|
| 31 |
|
|
|
|
| 32 |
## Ground Truth
|
| 33 |
|
| 34 |
Indoor ground-truth 6-DoF poses are recorded at **100 Hz** in a **10 m × 4 m** test area using **eight Qualisys Miqus M3 motion-capture cameras**.
|
|
|
|
| 1 |
---
|
| 2 |
+
viewer: false
|
| 3 |
license: gpl-3.0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4 |
---
|
| 5 |
|
| 6 |
# JEPLO Dataset
|
|
|
|
| 13 |
|
| 14 |
**Code:** [ASIG-X/JEPLO](https://github.com/ASIG-X/JEPLO)
|
| 15 |
|
| 16 |
+
This dataset can be used for evaluating legged odometry systems in challenging scenarios using proprioceptive and exteroceptive sensors.
|
| 17 |
+
|
| 18 |
## Dataset Overview
|
| 19 |
|
| 20 |
Publicly released to support the legged robotics community, the dataset comprises **eight sequences** recorded during our experiments: **five indoor sequences** and **three mixed indoor–outdoor sequences**.
|
|
|
|
| 26 |
- Joint encoder measurements.
|
| 27 |
- 6-DoF ground-truth trajectories.
|
| 28 |
|
| 29 |
+
|
| 30 |
## Ground Truth
|
| 31 |
|
| 32 |
Indoor ground-truth 6-DoF poses are recorded at **100 Hz** in a **10 m × 4 m** test area using **eight Qualisys Miqus M3 motion-capture cameras**.
|