Rename to pointnext-sm.scanobjectnn-hardest.openpoints

README.md

@@ -11,12 +11,12 @@ tags:
 datasets:
 - scanobjectnn
 model-index:
-- name: pointnext-sm.scanobjectnn.openpoints
+- name: pointnext-sm.scanobjectnn-hardest.openpoints
   results:
   - task:
       type: point-cloud-classification
     dataset:
-      name: ScanObjectNN
+      name: ScanObjectNN (PB_T50_RS)
       type: scanobjectnn
     metrics:
     - name: OA
@@ -27,9 +27,9 @@ model-index:
       value: 86.75
 ---
 
-# Model card for pointnext-sm.scanobjectnn.openpoints
+# Model card for pointnext-sm.scanobjectnn-hardest.openpoints
 
-A PointNeXt point cloud classification model (scaled PointNet++ with inverted residual blocks). Trained on ScanObjectNN.
+A PointNeXt point cloud classification model (scaled PointNet++ with inverted residual blocks). Trained on ScanObjectNN (PB_T50_RS).
 
 ## Model Details
 
@@ -39,7 +39,7 @@ A PointNeXt point cloud classification model (scaled PointNet++ with inverted re
   - Input channels: 4
   - Classes: 15
   - Features: 512
-- **Dataset:** ScanObjectNN
+- **Dataset:** ScanObjectNN (PB_T50_RS)
 - **Metrics:** OA 88.17, mAcc 86.75 (reference 88.20)
 - **Paper:** [PointNeXt: Revisiting PointNet++ with Improved Training and Scaling Strategies](https://arxiv.org/abs/2206.04670)
 - **Converted from:** [guochengqian/PointNeXt](https://github.com/guochengqian/PointNeXt) (MIT)
@@ -59,7 +59,7 @@ import torch_pointcloud as tp
 from torch_pointcloud.utils.data import collate
 
 model, info = tp.create_model(
-    "pointnext-sm.scanobjectnn.openpoints",
+    "pointnext-sm.scanobjectnn-hardest.openpoints",
     task="classification",
     pretrained=True,
     return_info=True,