Update RefRef_test.py

RefRef_test.py

@@ -1,56 +1,90 @@
-# RefRef_test.py
-from datasets import Dataset, DatasetDict, Features, Image, Value
 import json
 import os
-from PIL import Image as PILImage
+import datasets
 
-def load_RefRef_test(data_dir, config_name, splits=["train", "val", "test"]):
-    dataset_dict = {}
-    
-    # Define features to match the YAML
-    features = Features({
-        "image": Image(),
-        "depth": Image(),
-        "mask": Image(),
-        "transform_matrix": Value("float64", shape=(4, 4)),
-        "rotation": Value("float32")
-    })
-
-    print("1111!")
+_CITATION = """\
+@InProceedings{...},
+title = {Your Dataset Title},
+author={Your Name},
+year={2023}
+}
+"""
+
+_DESCRIPTION = """\
+Dataset containing multi-view images with camera poses, depth maps, and masks for NeRF training.
+"""
+
+_LICENSE = "MIT"
+
+class RefRef_test(datasets.GeneratorBasedBuilder):
+    """A dataset loader for NeRF-style data with camera poses, depth maps, and masks."""
+
+    VERSION = datasets.Version("1.0.0")
     
-    for split in splits:
-        json_path = os.path.join(data_dir, config_name, f"transforms_{split}.json")
+    # No multiple configs needed - using single configuration
+    BUILDER_CONFIGS = [
+        datasets.BuilderConfig(
+            name="default",
+            version=VERSION,
+            description="Default configuration for NeRF dataset"
+        )
+    ]
+
+    def _info(self):
+        features = datasets.Features({
+            "image": datasets.Image(),
+            "depth": datasets.Image(),
+            "mask": datasets.Image(),
+            "transform_matrix": datasets.Sequence(
+                datasets.Sequence(datasets.Value("float64"), length=4),
+                length=4
+            ),
+            "rotation": datasets.Value("float32")
+        })
         
-        with open(json_path, "r") as f:
-            data = json.load(f)
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=features,
+            homepage="",
+            license=_LICENSE,
+            citation=_CITATION
+        )
+
+    def _split_generators(self, dl_manager):
+        return [
+            datasets.SplitGenerator(
+                name=datasets.Split.TRAIN,
+                gen_kwargs={"split": "train"}
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split.VALIDATION,
+                gen_kwargs={"split": "val"}
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split.TEST,
+                gen_kwargs={"split": "test"}
+            ),
+        ]
+
+    def _generate_examples(self, split):
+        base_path = os.path.join("./")  # Update this path
         
-        examples = []
-        for frame in data["frames"]:
-            # Resolve relative paths to load images
-            base_dir = os.path.dirname(json_path)
-            
-            # Load image
-            image_path = os.path.join(base_dir, frame["file_path"])
-            image = PILImage.open(image_path)
+        # Assuming your directory structure has scene folders (ball, ampoule) 
+        # with transforms_{split}.json files
+        for scene in ["ball", "ampoule"]:  # Add all your scene names here
+            json_path = os.path.join(base_path, scene, f"transforms_{split}.json")
             
-            # Load depth
-            depth_path = os.path.join(base_dir, frame["depth_file_path"])
-            depth = PILImage.open(depth_path)
+            with open(json_path, "r") as f:
+                data = json.load(f)
             
-            # Load mask
-            mask_path = os.path.join(base_dir, frame["mask_file_path"])
-            mask = PILImage.open(mask_path)
-            
-            # Create example
-            example = {
-                "image": image,
-                "depth": depth,
-                "mask": mask,
-                "transform_matrix": frame["transform_matrix"],
-                "rotation": frame.get("rotation", 0.0)
-            }
-            examples.append(example)
-        
-        dataset_dict[split] = Dataset.from_list(examples, features=features)
-    
-    return DatasetDict(dataset_dict)
+            for idx, frame in enumerate(data["frames"]):
+                # Construct full paths relative to JSON file location
+                base_dir = os.path.dirname(json_path)
+                
+                yield f"{scene}_{split}_{idx}", {
+                    "image": os.path.join(base_dir, frame["file_path"]),
+                    "depth": os.path.join(base_dir, frame["depth_file_path"]),
+                    "mask": os.path.join(base_dir, frame["mask_file_path"]),
+                    "transform_matrix": frame["transform_matrix"],
+                    "rotation": frame.get("rotation", 0.0)
+                }