Update README.md

README.md

@@ -74,12 +74,52 @@ The dataset enables **vision-action pretraining** on large-scale human gameplay
 | VALORANT | 1 | **0.25 h (911.94 s)** | 911.94 s (15.20 min) |
 
 ## Usage Example
+
+```bash
+pip install mcap-owa-support owa-msgs huggingface_hub
+```
+
 ```python
-from datasets import load_dataset
+from huggingface_hub import hf_hub_download
+from mcap_owa.highlevel import OWAMcapReader
+
+# Download MCAP and video files
+mcap_file = hf_hub_download(
+    repo_id="open-world-agents/D2E-480p",
+    filename="Apex_Legends/0805_01.mcap",
+    repo_type="dataset"
+)
+video_file = hf_hub_download(
+    repo_id="open-world-agents/D2E-480p",
+    filename="Apex_Legends/0805_01.mkv",
+    repo_type="dataset"
+)
 
-dataset = load_dataset("open-world-agents/D2E", split="train")
+# Load and iterate through messages
+with OWAMcapReader(mcap_file) as reader:
+    # Screen frames with video loading
+    for msg in reader.iter_messages(topics=["screen"]):
+        screen = msg.decoded
+        screen.resolve_relative_path(mcap_file)  # Resolve video path
+        frame = screen.load_frame_array()  # Load actual frame from video
+        print(f"Frame: {frame.shape}")
+        break
+
+    # Keyboard events
+    for msg in reader.iter_messages(topics=["keyboard"]):
+        kbd = msg.decoded
+        print(f"Keyboard: {kbd.event_type} VK={kbd.vk}")
+        break
+
+    # Raw mouse events (we use this instead of "mouse", see why: https://open-world-agents.github.io/open-world-agents/env/plugins/desktop/#raw-mouse-input)
+    for msg in reader.iter_messages(topics=["mouse/raw"]):
+        mouse = msg.decoded
+        print(f"Mouse: dx={mouse.last_x} dy={mouse.last_y}")
+        break
 ```
 
+For more details, see the [OWA Documentation](https://open-world-agents.github.io/open-world-agents/).
+
 ## Citation
 If you find this work useful, please cite our paper:
 ```