Update README.md

README.md

@@ -102,22 +102,23 @@ instances["categories"]
 ## Build the dataset and upload to Hub
 
 ```py
-!wget http://images.cocodataset.org/zips/train2017.zip
-!wget http://images.cocodataset.org/zips/val2017.zip
-!wget http://images.cocodataset.org/annotations/annotations_trainval2017.zip
+!pip install -U -q datasets
 
-!unzip train2017.zip
-!unzip val2017.zip
-!unzip annotations_trainval2017.zip
+# Download and unzip COCO 2017
+!wget -q http://images.cocodataset.org/zips/train2017.zip
+!wget -q http://images.cocodataset.org/zips/val2017.zip
+!wget -q http://images.cocodataset.org/annotations/annotations_trainval2017.zip
+
+!unzip -q train2017.zip
+!unzip -q val2017.zip
+!unzip -q annotations_trainval2017.zip
 
 import json
+import shutil
 from pathlib import Path
 from tqdm import tqdm
-from huggingface_hub import upload_folder
-from datasets import Dataset, DatasetDict, Features, Value, Sequence, Array2D
-import shutil
+from datasets import load_dataset
 
-# === Paths ===
 base_dir = Path("/content")
 splits = {
     "train": {
@@ -130,11 +131,15 @@ splits = {
     }
 }
 output_dir = base_dir / "coco_imagefolder"
+output_dir.mkdir(parents=True, exist_ok=True)
 
-# Clean existing directory
-if output_dir.exists():
-    shutil.rmtree(output_dir)
-output_dir.mkdir(parents=True)
+def normalize_segmentation(segmentation):
+    if isinstance(segmentation, list):
+        if all(isinstance(poly, list) for poly in segmentation):
+            return segmentation  # already a list of polygons
+        elif all(isinstance(pt, (int, float)) for pt in segmentation):
+            return [segmentation]  # wrap single polygon
+    return []  # skip RLE or malformed segmentations
 
 def convert_coco_to_jsonl(image_dir, annotation_path, output_metadata_path):
     with open(annotation_path) as f:
@@ -145,42 +150,46 @@ def convert_coco_to_jsonl(image_dir, annotation_path, output_metadata_path):
 
     for ann in data['annotations']:
         img_id = ann['image_id']
-        bbox = ann['bbox']  # [x, y, width, height]
+        bbox = ann['bbox']
         category = ann['category_id']
+        segmentation = normalize_segmentation(ann['segmentation'])
+
+        if not segmentation:
+            continue  # skip if malformed or RLE
 
         if img_id not in annotations_by_image:
             annotations_by_image[img_id] = {
                 "file_name": id_to_filename[img_id],
                 "objects": {
                     "bbox": [],
-                    "categories": []
+                    "segmentation": [],
+                    "categories": [],
                 }
             }
 
         annotations_by_image[img_id]["objects"]["bbox"].append(bbox)
+        annotations_by_image[img_id]["objects"]["segmentation"].append(segmentation)
         annotations_by_image[img_id]["objects"]["categories"].append(category)
 
     with open(output_metadata_path, "w") as f:
-        for img_id, metadata in annotations_by_image.items():
+        for metadata in annotations_by_image.values():
             json.dump(metadata, f)
             f.write("\n")
 
-# Convert and copy files to imagefolder-style structure
+# Build imagefolder structure
 for split, info in splits.items():
     split_dir = output_dir / split
-    split_dir.mkdir(parents=True)
+    split_dir.mkdir(parents=True, exist_ok=True)
 
     # Copy images
     for img_path in tqdm(info["image_dir"].glob("*.jpg"), desc=f"Copying {split} images"):
         shutil.copy(img_path, split_dir / img_path.name)
 
-    # Convert annotations
+    # Write JSONL metadata
     metadata_path = split_dir / "metadata.jsonl"
     convert_coco_to_jsonl(split_dir, info["annotation_file"], metadata_path)
 
-# HF Dataset
-from datasets import load_dataset
-
-dataset = load_dataset("imagefolder", data_dir="/content/coco_imagefolder")
+# Load and push
+dataset = load_dataset("imagefolder", data_dir=str(output_dir))
 dataset.push_to_hub("ariG23498/coco2017")
 ```