initial

.gitattributes

@@ -0,0 +1,2 @@
+mapillary_dataset/**/*.jpg filter=lfs diff=lfs merge=lfs -text
+main_test_samples.csv filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,3 @@
+---
+license: cc-by-nc-4.0
+---

geochain.py

@@ -0,0 +1,149 @@
+import csv
+import os
+import datasets
+from PIL import Image 
+from pathlib import Path 
+from typing import Union, Optional 
+
+def crop_bottom_rows(image_path: Union[str, Path], num_rows_to_crop: int) -> Optional[Image.Image]:
+    """
+    Reads an image from a file path and crops the bottom few rows.
+
+    Args:
+        image_path: The path to the image file.
+        num_rows_to_crop: The number of rows to crop from the bottom.
+
+    Returns:
+        The cropped image object (Pillow.Image.Image), or None if an error occurs.
+    """
+    try:
+        img = Image.open(image_path)
+        width, height = img.size
+
+        if not isinstance(num_rows_to_crop, int) or num_rows_to_crop < 0:
+            print(f"Error: Number of rows to crop ({num_rows_to_crop}) must be a non-negative integer for {image_path}.")
+            return None
+        if num_rows_to_crop >= height:
+            # If num_rows_to_crop equals height, it results in a 0-height image, which is usually problematic.
+            # If it's greater, it's definitely an issue.
+            print(f"Error: Number of rows to crop ({num_rows_to_crop}) is >= image height ({height}) for {image_path}. Cropping would result in an empty or invalid image.")
+            return None # Return None if crop results in zero or negative height
+
+        left, upper, right, lower = 0, 0, width, height - num_rows_to_crop
+        
+        # Ensure lower is not less than upper (which is 0)
+        if lower <= upper:
+            print(f"Warning: Calculated crop area for {image_path} results in zero or negative height (lower: {lower}, upper: {upper}). Returning None.")
+            return None
+
+        cropped_img = img.crop((left, upper, right, lower))
+        return cropped_img
+    except FileNotFoundError:
+        print(f"Error: Image file not found at {image_path}")
+        return None
+    except Exception as e:
+        print(f"An unexpected error occurred while cropping {image_path}: {e}")
+        return None
+
+
+
+_DESCRIPTION = "Geochain dataset with a mini_test split (cropped images+metadata) and a test split (metadata only)."
+_HOMEPAGE = "https://huggingface.co/datasets/sahitiy51/geochain"
+
+class GeochainDataset(datasets.GeneratorBasedBuilder):
+    """Geochain dataset with mini_test (cropped images) and test (CSV) splits."""
+
+    VERSION = datasets.Version("1.0.0") 
+
+    def _info(self):
+        features = datasets.Features({
+            "image": datasets.Image(), 
+            "key": datasets.Value("string"),
+            "sub_folder": datasets.Value("string"),
+            "locatability_score": datasets.Value("float"),
+            "lat": datasets.Value("float"),
+            "lon": datasets.Value("float"),
+            "city": datasets.Value("string"),
+            "class_mapping": datasets.Value("string"),
+            "sequence_key": datasets.Value("string"),
+        })
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=features,
+            homepage=_HOMEPAGE,
+        )
+
+    def _split_generators(self, dl_manager):
+        path_to_mini_test_csv = os.path.join(dl_manager.manual_dir, "test_samples.csv")
+        path_to_test_csv = os.path.join(dl_manager.manual_dir, "main_test_samples.csv")
+        path_to_original_images_base = os.path.join(dl_manager.manual_dir, "mapillary_dataset")
+
+        return [
+            datasets.SplitGenerator(
+                name="mini_test",
+                gen_kwargs={
+                    "csv_path": path_to_mini_test_csv,
+                    "original_images_base_dir": path_to_original_images_base,
+                    "split_name": "mini_test",
+                },
+            ),
+            datasets.SplitGenerator(
+                name="test",
+                gen_kwargs={
+                    "csv_path": path_to_test_csv,
+                    "original_images_base_dir": None,
+                    "split_name": "test",
+                },
+            ),
+        ]
+
+    def _generate_examples(self, csv_path, original_images_base_dir, split_name):
+        with open(csv_path, encoding="utf-8") as f:
+            reader = csv.DictReader(f)
+            for idx, row in enumerate(reader):
+                example_id = f"{split_name}_{idx}"
+
+                if split_name == "mini_test":
+                    image_key_from_csv = row["key"]
+                    city_for_path = row["city"]
+                    sub_folder_for_path = row["sub_folder"]
+
+                    full_original_image_path = os.path.join(
+                        original_images_base_dir,
+                        city_for_path,
+                        sub_folder_for_path,
+                        "images",
+                        f"{image_key_from_csv}.jpg"  # Assuming .jpg extension for originals
+                    )
+
+                    # --- Apply Your Cropping Logic ---
+                    # !!! IMPORTANT: Set your desired number of rows to crop here !!!
+                    NUM_ROWS_TO_CROP_VALUE = 30 # EXAMPLE VALUE - CHANGE THIS
+                    
+                    cropped_image_data = crop_bottom_rows(full_original_image_path, NUM_ROWS_TO_CROP_VALUE)
+                    # cropped_image_data will be a PIL.Image object or None if cropping failed
+                    # --- End of Cropping Logic ---
+
+                    yield example_id, {
+                        "image": cropped_image_data, 
+                        "key": None, 
+                        "sub_folder": None,
+                        "locatability_score": float(row["locatability_score"]) if row["locatability_score"] and row["locatability_score"].strip() else None,
+                        "lat": float(row["lat"]) if row["lat"] and row["lat"].strip() else None,
+                        "lon": float(row["lon"]) if row["lon"] and row["lon"].strip() else None,
+                        "city": None, 
+                        "class_mapping": row["class_mapping"],
+                        "sequence_key": row["sequence_key"],
+                    }
+                elif split_name == "test":
+                    yield example_id, {
+                        "image": None,
+                        "key": row["key"],
+                        "sub_folder": row["sub_folder"],
+                        "locatability_score": float(row["locatability_score"]) if row["locatability_score"] and row["locatability_score"].strip() else None,
+                        "lat": float(row["lat"]) if row["lat"] and row["lat"].strip() else None,
+                        "lon": float(row["lon"]) if row["lon"] and row["lon"].strip() else None,
+                        "city": row["city"],
+                        "class_mapping": row["class_mapping"],
+                        "sequence_key": row["sequence_key"],
+                    }

main_test_samples.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0eeb2081c362c49a1f32bc7f1f597da3efc780c4da807bb806ba8acc5d14974b
+size 632631444