Refactor

detect-objects.py

@@ -67,8 +67,8 @@ from transformers import Sam3Processor, Sam3Model
 # Configure logging
 logging.basicConfig(
     level=logging.INFO,
-    format='%(asctime)s - %(levelname)s - %(message)s',
-    datefmt='%H:%M:%S'
+    format="%(asctime)s - %(levelname)s - %(message)s",
+    datefmt="%H:%M:%S",
 )
 logger = logging.getLogger(__name__)
 
@@ -85,59 +85,53 @@ def parse_args():
     parser = argparse.ArgumentParser(
         description="Detect objects in images using SAM3",
         formatter_class=argparse.RawDescriptionHelpFormatter,
-        epilog=__doc__
+        epilog=__doc__,
     )
 
     # Required arguments
     parser.add_argument(
-        "input_dataset",
-        help="Input HuggingFace dataset ID (e.g., 'username/dataset')"
+        "input_dataset", help="Input HuggingFace dataset ID (e.g., 'username/dataset')"
     )
     parser.add_argument(
-        "output_dataset",
-        help="Output HuggingFace dataset ID (e.g., 'username/output')"
+        "output_dataset", help="Output HuggingFace dataset ID (e.g., 'username/output')"
     )
 
     # Object detection configuration
     parser.add_argument(
         "--classes",
         required=True,
-        help="Comma-separated list of object classes to detect (e.g., 'photograph,illustration,diagram')"
+        help="Comma-separated list of object classes to detect (e.g., 'photograph,illustration,diagram')",
     )
     parser.add_argument(
         "--confidence-threshold",
         type=float,
         default=0.5,
-        help="Minimum confidence score for detections (default: 0.5)"
+        help="Minimum confidence score for detections (default: 0.5)",
     )
     parser.add_argument(
         "--mask-threshold",
         type=float,
         default=0.5,
-        help="Threshold for mask generation (default: 0.5)"
+        help="Threshold for mask generation (default: 0.5)",
     )
 
     # Dataset configuration
     parser.add_argument(
         "--image-column",
         default="image",
-        help="Name of the column containing images (default: 'image')"
+        help="Name of the column containing images (default: 'image')",
     )
     parser.add_argument(
-        "--split",
-        default="train",
-        help="Dataset split to process (default: 'train')"
+        "--split", default="train", help="Dataset split to process (default: 'train')"
     )
     parser.add_argument(
         "--max-samples",
         type=int,
         default=None,
-        help="Maximum number of samples to process (for testing)"
+        help="Maximum number of samples to process (for testing)",
     )
     parser.add_argument(
-        "--shuffle",
-        action="store_true",
-        help="Shuffle dataset before processing"
+        "--shuffle", action="store_true", help="Shuffle dataset before processing"
     )
 
     # Processing configuration
@@ -145,30 +139,28 @@ def parse_args():
         "--batch-size",
         type=int,
         default=4,
-        help="Batch size for processing (default: 4)"
+        help="Batch size for processing (default: 4)",
     )
     parser.add_argument(
         "--model",
         default="facebook/sam3",
-        help="SAM3 model ID (default: 'facebook/sam3')"
+        help="SAM3 model ID (default: 'facebook/sam3')",
     )
     parser.add_argument(
         "--dtype",
         default="bfloat16",
         choices=["float32", "float16", "bfloat16"],
-        help="Model precision (default: 'bfloat16')"
+        help="Model precision (default: 'bfloat16')",
     )
 
     # Output configuration
     parser.add_argument(
-        "--private",
-        action="store_true",
-        help="Make output dataset private"
+        "--private", action="store_true", help="Make output dataset private"
     )
     parser.add_argument(
         "--hf-token",
         default=None,
-        help="HuggingFace token (default: uses HF_TOKEN env var or cached token)"
+        help="HuggingFace token (default: uses HF_TOKEN env var or cached token)",
     )
 
     return parser.parse_args()
@@ -179,8 +171,8 @@ def load_and_validate_dataset(
     split: str,
     image_column: str,
     max_samples: int = None,
-    shuffle: bool = False, 
-    hf_token: str = None
+    shuffle: bool = False,
+    hf_token: str = None,
 ) -> Dataset:
     """Load dataset and validate it has the required image column."""
     logger.info(f"📂 Loading dataset: {dataset_id} (split: {split})")
@@ -218,7 +210,7 @@ def process_batch(
     processor: Sam3Processor,
     model: Sam3Model,
     confidence_threshold: float,
-    mask_threshold: float
+    mask_threshold: float,
 ) -> Dict[str, List[List[Dict[str, Any]]]]:
     """Process a batch of images and return detections."""
     images = batch[image_column]
@@ -228,9 +220,7 @@ def process_batch(
     for img in images:
         if isinstance(img, str):
             img = Image.open(img)
-        if img.mode == "L":
-            img = img.convert("RGB")
-        elif img.mode != "RGB":
+        if img.mode == "L" or img.mode != "RGB":
             img = img.convert("RGB")
         pil_images.append(img)
 
@@ -242,7 +232,7 @@ def process_batch(
         inputs = processor(
             images=pil_images,
             text=class_names,  # All class names as prompts
-            return_tensors="pt"
+            return_tensors="pt",
         ).to(model.device)
 
         with torch.no_grad():
@@ -253,7 +243,7 @@ def process_batch(
             outputs,
             threshold=confidence_threshold,
             mask_threshold=mask_threshold,
-            target_sizes=original_sizes
+            target_sizes=original_sizes,
         )
 
     except Exception as e:
@@ -264,9 +254,9 @@ def process_batch(
     # Convert to HuggingFace object detection format
     batch_objects = []
     for result in results:
-        boxes = result.get('boxes', torch.tensor([]))
-        scores = result.get('scores', torch.tensor([]))
-        labels = result.get('labels', torch.tensor([]))
+        boxes = result.get("boxes", torch.tensor([]))
+        scores = result.get("scores", torch.tensor([]))
+        labels = result.get("labels", torch.tensor([]))
 
         # Handle empty results
         if len(boxes) == 0:
@@ -276,9 +266,7 @@ def process_batch(
         # Build list of detections
         detections = []
         for box, score, label_idx in zip(
-            boxes.cpu().numpy(),
-            scores.cpu().numpy(),
-            labels.cpu().numpy()
+            boxes.cpu().numpy(), scores.cpu().numpy(), labels.cpu().numpy()
         ):
             x1, y1, x2, y2 = box
             width = x2 - x1
@@ -287,7 +275,7 @@ def process_batch(
             detection = {
                 "bbox": [float(x1), float(y1), float(width), float(height)],
                 "category": int(label_idx),  # Index into class_names
-                "score": float(score)
+                "score": float(score),
             }
             detections.append(detection)
 
@@ -298,11 +286,12 @@ def process_batch(
 
 def main():
     args = parse_args()
-
     # Parse class names
-    class_names = [name.strip() for name in args.classes.split(',')]
-    if not class_names or any(not name for name in class_names):
-        logger.error("❌ Invalid --classes argument. Provide comma-separated class names.")
+    class_names = [name.strip() for name in args.classes.split(",")]
+    if not class_names or not all(class_names):
+        logger.error(
+            "❌ Invalid --classes argument. Provide comma-separated class names."
+        )
         sys.exit(1)
 
     logger.info("🚀 SAM3 Object Detection")
@@ -325,7 +314,7 @@ def main():
         args.image_column,
         args.max_samples,
         args.shuffle,
-        args.hf_token
+        args.hf_token,
     )
 
     # Load model
@@ -333,9 +322,7 @@ def main():
     try:
         processor = Sam3Processor.from_pretrained(args.model)
         model = Sam3Model.from_pretrained(
-            args.model,
-            torch_dtype=getattr(torch, args.dtype),
-            device_map="auto"
+            args.model, torch_dtype=getattr(torch, args.dtype), device_map="auto"
         )
         logger.info(f"✅ Model loaded on {model.device}")
     except Exception as e:
@@ -353,42 +340,47 @@ def main():
             processor,
             model,
             args.confidence_threshold,
-            args.mask_threshold
+            args.mask_threshold,
         ),
         batched=True,
         batch_size=args.batch_size,
-        desc="Detecting objects"
+        desc="Detecting objects",
     )
 
     # Create dynamic features with ClassLabel
     logger.info("📊 Creating output schema...")
     new_features = processed_dataset.features.copy()
-    new_features["objects"] = Sequence({
-        "bbox": Sequence(Value("float32"), length=4),
-        "category": ClassLabel(names=class_names),
-        "score": Value("float32")
-    })
+    new_features["objects"] = Sequence(
+        {
+            "bbox": Sequence(Value("float32"), length=4),
+            "category": ClassLabel(names=class_names),
+            "score": Value("float32"),
+        }
+    )
 
     # Cast to proper types
     processed_dataset = processed_dataset.cast(new_features)
 
     # Calculate statistics
     total_detections = sum(len(objs) for objs in processed_dataset["objects"])
-    images_with_detections = sum(1 for objs in processed_dataset["objects"] if len(objs) > 0)
+    images_with_detections = sum(len(objs) > 0 for objs in processed_dataset["objects"])
 
     logger.info("✅ Detection complete!")
     logger.info(f"   Total detections: {total_detections}")
-    logger.info(f"   Images with detections: {images_with_detections}/{len(processed_dataset)}")
-    logger.info(f"   Average detections per image: {total_detections/len(processed_dataset):.2f}")
+    logger.info(
+        f"   Images with detections: {images_with_detections}/{len(processed_dataset)}"
+    )
+    logger.info(
+        f"   Average detections per image: {total_detections / len(processed_dataset):.2f}"
+    )
 
     # Push to hub
     logger.info(f"📤 Pushing to HuggingFace Hub: {args.output_dataset}")
     try:
-        processed_dataset.push_to_hub(
-            args.output_dataset,
-            private=args.private
+        processed_dataset.push_to_hub(args.output_dataset, private=args.private)
+        logger.info(
+            f"✅ Dataset available at: https://huggingface.co/datasets/{args.output_dataset}"
         )
-        logger.info(f"✅ Dataset available at: https://huggingface.co/datasets/{args.output_dataset}")
     except Exception as e:
         logger.error(f"❌ Failed to push to hub: {e}")
         logger.info("💾 Saving locally as backup...")