DenisaBumba
/

rfdetr-instance-segmentation-leibniz-manuscripts-v2

Image Segmentation

object-detection

instance-segmentation

layout-analysis

document-analysis

historical-documents

Model card Files Files and versions

DenisaBumba commited on 8 days ago

Commit

eddbcdc

·

verified ·

1 Parent(s): e153070

Update README.md

Files changed (1) hide show

README.md +3 -3

README.md CHANGED Viewed

@@ -59,8 +59,8 @@ Results on the validation set.
 | Checkpoint | Epoch | segm mAP@50:95 |
 |---|---|---|
-| `checkpoint_best_regular.pth` | 40/60 | 0.4670 |
-| `checkpoint_best_ema.pth` | 40/60 | **0.4700** |
 Per-class validation results for EMA checkpoint **v2**.
@@ -81,7 +81,7 @@ Per-class validation results for EMA checkpoint **v2**.
 from rfdetr import RFDETRSegNano
 import cv2
 from rfdetr.assets.coco_classes import COCO_CLASSES
-model = RFDETRSegNano(pretrain_weights="checkpoint_best_regular.pth")
 image = cv2.imread("image.jpg")
 detections = model.predict(image, threshold=0.4)
 labels = [f"{COCO_CLASSES[class_id]}" for class_id in detections.class_id]

 | Checkpoint | Epoch | segm mAP@50:95 |
 |---|---|---|
+| `checkpoint_best_regular_v2.pth` | 40/60 | 0.4670 |
+| `checkpoint_best_ema_v2.pth` | 40/60 | **0.4700** |
 Per-class validation results for EMA checkpoint **v2**.
 from rfdetr import RFDETRSegNano
 import cv2
 from rfdetr.assets.coco_classes import COCO_CLASSES
+model = RFDETRSegNano(pretrain_weights="checkpoint_best_regular_v2.pth")
 image = cv2.imread("image.jpg")
 detections = model.predict(image, threshold=0.4)
 labels = [f"{COCO_CLASSES[class_id]}" for class_id in detections.class_id]