Update README.md
Browse files
README.md
CHANGED
|
@@ -13,6 +13,7 @@ task_categories:
|
|
| 13 |
|
| 14 |
This dataset is designed for **training and evaluating models on Vision Language Navigation task**. It is collected using unitree go2 navigating through indoor environments from the **Matterport dataset** (provided by VLN-CE-Isaac benchmark), rendered with **Isaac Lab**. Each episode in the dataset represents a navigation task where the robot follows an **expert path** using a **Proportional-Derivative (PD) controller**, capturing **visual observations** (RGB images and depth maps) and **discrete actions** at regular intervals (40 simulation steps). The dataset is intended to support research in areas such as imitation learning, reinforcement learning, and vision-language navigation.
|
| 15 |
|
|
|
|
| 16 |
---
|
| 17 |
|
| 18 |
## Dataset Structure
|
|
|
|
| 13 |
|
| 14 |
This dataset is designed for **training and evaluating models on Vision Language Navigation task**. It is collected using unitree go2 navigating through indoor environments from the **Matterport dataset** (provided by VLN-CE-Isaac benchmark), rendered with **Isaac Lab**. Each episode in the dataset represents a navigation task where the robot follows an **expert path** using a **Proportional-Derivative (PD) controller**, capturing **visual observations** (RGB images and depth maps) and **discrete actions** at regular intervals (40 simulation steps). The dataset is intended to support research in areas such as imitation learning, reinforcement learning, and vision-language navigation.
|
| 15 |
|
| 16 |
+
Download Link: https://1drv.ms/u/c/7c29b3d8450f09f4/EXYjBta3ANVCpKyTY6gwl5gBrRVlxTkqo64CX_-fQZBPxw?e=HCZBlI
|
| 17 |
---
|
| 18 |
|
| 19 |
## Dataset Structure
|