Upload prediction_example.py with huggingface_hub

prediction_example.py

@@ -0,0 +1,52 @@
+import os
+from pathlib import Path
+from PIL import Image
+import pandas as pd
+
+import Models
+
+config = {
+    "model_root": "models",
+    "hf_model_repo": "SurfaceAI/models",
+    "models": {
+        "surface_type": "v1/surface_type_v1.pt",
+        "surface_quality": {
+            "asphalt": "v1/surface_quality_asphalt_v1.pt",
+            "concrete": "v1/surface_quality_concrete_v1.pt",
+            "paving_stones": "v1/surface_quality_paving_stones_v1.pt",
+            "sett": "v1/surface_quality_sett_v1.pt",
+            "unpaved": "v1/surface_quality_unpaved_v1.pt"
+        },
+        "road_type": "v1/road_type_v1.pt"
+    },
+    "gpu_kernel": 0,
+    "transform_surface": {
+        "resize": 384,
+        "crop": "lower_middle_half"
+    },
+    "transform_road_type": {
+        "resize": 384,
+        "crop": "lower_half"
+    },
+}
+
+root_path = Path(os.path.abspath(__file__)).parent
+
+image_ids = [
+    "1351262795801113",
+    "153111940043147",
+    "1424818291203908",
+]
+
+image_data = []
+for id in image_ids:
+    path = root_path / "example_images" / f"{id}.jpg"
+    try:
+        image_data.append(Image.open(path))
+    except Exception as e:
+        print(f'{e}: Not found or corrupted image: {path}')
+        continue
+
+md = Models.ModelInterface(config=config)
+df = md.batch_classifications(image_data, image_ids)
+df.to_csv("example_prediction.csv")