nvidia
/

Cosmos-Transfer-LidarGen

Model card Files Files and versions

ShengyHuang commited on Oct 23, 2025

Commit

25fd741

·

verified ·

1 Parent(s): af63509

Update README.md

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -43,7 +43,7 @@ This model was developed based on Cosmos-Predict1 Single2MultiView foundation mo
 **Input Type(s):** Image <br>
 **Input Format:** Red, Green, Blue (RGB) <br>
 **Input Parameters:** Two-Dimensional (2D) <br>
-**Other Properties Related to Input:** Multi-view RGB images from 3 f-theta cameras (front wide 120° FOV, rear left 70° FOV, rear right 70° FOV). Input resolution of [720, 1280] cropped to [704, 1280]. Synchronized camera views required. <br>
 ## Output: <br>
 **Output Type(s):** Lidar <br>

 **Input Type(s):** Image <br>
 **Input Format:** Red, Green, Blue (RGB) <br>
 **Input Parameters:** Two-Dimensional (2D) <br>
+**Other Properties Related to Input:** Multi-view RGB images from 3 f-theta cameras (front wide 120° FOV, rear left 70° FOV, rear right 70° FOV). Input resolution of [720, 1280] cropped to [704, 1280]. Synchronized camera views required. Note that this model only works for Nvidia camera-lidar rig configuration. Users will need to post-train their own LidarGen model following the provided [receipe](https://github.com/nv-tlabs/Cosmos-Drive-Dreams/blob/main/cosmos-transfer-lidargen/examples/lidar_diffusion.md) on different data. <br>
 ## Output: <br>
 **Output Type(s):** Lidar <br>