Add paper, project and code links; add license and robotics metadata

README.md

@@ -1,24 +1,29 @@
 ---
 language:
-  - en
+- en
+license: mit
+size_categories:
+- n<1K
 task_categories:
-  - robotics
-  - reinforcement-learning
+- robotics
+- reinforcement-learning
+pretty_name: World Model Robot Manipulation Dataset (Our-50)
 tags:
-  - robotics
-  - manipulation
-  - imitation-learning
-  - world-model
-  - robot-learning
-  - tabletop
-pretty_name: "World Model Robot Manipulation Dataset (Our-50)"
-size_categories:
-  - n<1K
+- robotics
+- manipulation
+- imitation-learning
+- world-model
+- robot-learning
+- tabletop
 ---
 
 # World Model Robot Manipulation Dataset
 
-A dataset of real-robot tabletop manipulation trajectories collected for world model training and imitation learning research. The setup follows DROID Dataset. Each trajectory pairs multi-camera video, proprioceptive state/action sequences, natural language task descriptions, and dense reward annotations with pre-extracted visual latents.
+[**Project Page**](https://arnavkj1995.github.io/WEAVER/) | [**Paper**](https://huggingface.co/papers/2606.13672) | [**Code**](https://github.com/arnavkj1995/WEAVER)
+
+A dataset of real-robot tabletop manipulation trajectories collected for world model training and imitation learning research. The setup follows the DROID Dataset. Each trajectory pairs multi-camera video, proprioceptive state/action sequences, natural language task descriptions, and dense reward annotations with pre-extracted visual latents.
+
+This dataset was introduced in the paper: **WEAVER, Better, Faster, Longer: An Effective World Model for Robotic Manipulation**.
 
 ## Dataset Summary
 
@@ -42,6 +47,60 @@ Five tabletop manipulation tasks, 50 train / 20 val trajectories per task.
 
 Each task has multiple natural-language paraphrases (e.g. *"put the marker in the cup"*, *"put the marker in the mug"*, *"pick up the marker and place it in the cup"*).
 
+## Usage Example
+
+### Download Dataset
+
+You can download the dataset using the `huggingface_hub` library:
+
+```python
+from huggingface_hub import snapshot_download
+
+local_dir = snapshot_download(
+    repo_id="yilin-wu/droid_ood_data",
+    repo_type="dataset",
+)
+```
+
+To download only annotations and metadata (without videos/latents):
+
+```python
+from huggingface_hub import snapshot_download
+
+local_dir = snapshot_download(
+    repo_id="yilin-wu/droid_ood_data",
+    repo_type="dataset",
+    allow_patterns=["annotations/**", "annotation_rewards/**", "norm_stats*.json"],
+)
+```
+
+### Loading Trajectories
+
+```python
+import json
+import numpy as np
+
+# Load a trajectory
+with open("annotations/train/0.json") as f:
+    traj = json.load(f)
+
+print(traj["texts"])          # ['pick up the bag of chips and place it on the green plate']
+print(traj["success"])        # 1
+print(traj["video_length"])   # e.g. 112
+
+# Joint positions: shape (T, 7)
+joint_pos = np.array(traj["observation.state.joint_position"])
+
+# Actions: shape (T, 7)
+actions = np.array(traj["action.joint_position"])
+
+# Visual latents: shape (3, T, 60, 256)
+lat = np.load(traj["latent_path"].replace("latents/", "latents/"))["latents"]
+
+# Rewards: shape (T,)
+rewards = np.array(traj["reward_progress"])
+```
+
 ## Data Structure
 
 ```
@@ -64,8 +123,6 @@ world_model_data_our_50/
 
 ### Annotation JSON Schema
 
-Each `.json` file contains one trajectory with the following fields:
-
 | Field | Type | Description |
 |-------|------|-------------|
 | `episode_id` | int | Sequential trajectory index within the split |
@@ -109,39 +166,20 @@ shape: (3, T, 60, 256)
         └─ cameras
 ```
 
-### Normalization Statistics
-
-`norm_stats_recorded.json` and `norm_stats_relabel.json` provide mean/std statistics for the `state` and `actions` modalities, suitable for normalizing inputs during training.
-
 ## Robot Setup
 
 - **Robot**: Franka Emika Robot arm with parallel-jaw gripper (Robotiq Gripper)
 - **Cameras**: 3 fixed cameras providing left, right, and wrist views
 - **Control frequency**: 5 Hz (matches video frame rate)
 
-## Usage Example
+## Citation
 
-```python
-import json
-import numpy as np
-
-# Load a trajectory
-with open("annotations/train/0.json") as f:
-    traj = json.load(f)
-
-print(traj["texts"])          # ['pick up the bag of chips and place it on the green plate']
-print(traj["success"])        # 1
-print(traj["video_length"])   # e.g. 112
-
-# Joint positions: shape (T, 7)
-joint_pos = np.array(traj["observation.state.joint_position"])
-
-# Actions: shape (T, 7)
-actions = np.array(traj["action.joint_position"])
-
-# Visual latents: shape (3, T, 60, 256)
-lat = np.load(traj["latent_path"].replace("latents/", "latents/"))["latents"]
-
-# Rewards: shape (T,)
-rewards = np.array(traj["reward_progress"])
+```bibtex
+@article{jain2026weaver,
+  title={WEAVER: Efficient World Models for Robot Video Prediction},
+  author={Arnav Kumar Jain and Yilin Wu and Jesse Farebrother and Gokul Swamy and Andrea Bajcsy},
+  journal={CoRR},
+  volume={abs/2606.13672},
+  year={2026}
+}
 ```