Upload 2 files

checkpoint_best_total.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:98f3d25a38ae7f9e598d02bead7903179ed99e8114ab99a31dfe9a81b69f532e
+size 127634110

script.py

@@ -0,0 +1,65 @@
+import os
+import torch
+import pandas as pd
+from PIL import Image
+import numpy as np
+from rfdetr import RFDETRBase
+
+
+def run_inference(model, image_path, conf_threshold, save_path):
+
+    test_images = sorted(os.listdir(image_path))
+
+    bboxes = []
+    category_ids = []
+    test_images_names = []
+
+    for image_name in test_images:
+        test_images_names.append(image_name)
+
+        image_file = os.path.join(image_path, image_name)
+        image = Image.open(image_file).convert("RGB")
+
+        preds = model.predict(image)
+
+        image_bboxes = []
+        image_categories = []
+
+        for box, score, label in zip(
+            preds["boxes"], preds["scores"], preds["labels"]
+        ):
+            if score >= conf_threshold:
+                xmin, ymin, xmax, ymax = box.tolist()
+                width = xmax - xmin
+                height = ymax - ymin
+
+                image_bboxes.append([xmin, ymin, width, height])
+                image_categories.append(int(label))
+
+        bboxes.append(image_bboxes)
+        category_ids.append(image_categories)
+
+    df_predictions = pd.DataFrame(columns=["file_name", "bbox", "category_id"])
+
+    for i in range(len(test_images_names)):
+        df_predictions.loc[i] = [
+            test_images_names[i],
+            str(bboxes[i]),
+            str(category_ids[i]),
+        ]
+
+    df_predictions.to_csv(save_path, index=False)
+
+
+if __name__ == "__main__":
+
+    TEST_IMAGE_PATH = "/tmp/data/test_images"
+    SUBMISSION_SAVE_PATH = "submission.csv"
+    CONF_THRESHOLD = 0.30
+
+    model = RFDETRBase(
+        checkpoint_path="checkpoint_best_ema.pth",
+        device="cuda" if torch.cuda.is_available() else "cpu"
+    )
+
+    run_inference(model, TEST_IMAGE_PATH, CONF_THRESHOLD, SUBMISSION_SAVE_PATH)