docs: improve usage examples and clarify output formats

- Add output format comments to pipeline example
- Note that pipeline auto-manages GPU and returns original resolution
- Update production example with auto GPU detection
- Clarify preprocessing difference (cv2.INTER_AREA vs PIL LANCZOS)

README.md

@@ -34,10 +34,13 @@ This model segments human images into 18 semantic categories including body part
 ```python
 from transformers import pipeline
 
-parser = pipeline("image-segmentation", model="fashn-ai/fashn-human-parser")
-result = parser("path/to/image.jpg")
+pipe = pipeline("image-segmentation", model="fashn-ai/fashn-human-parser")
+result = pipe("image.jpg")
+# result is a list of dicts with 'label', 'score', 'mask' for each detected class
 ```
 
+The pipeline automatically manages GPU/CPU and returns per-class masks at the original image resolution.
+
 ### Explicit Usage
 
 ```python
@@ -67,7 +70,7 @@ predictions = upsampled.argmax(dim=1).squeeze().numpy()
 
 ### Production Usage (Recommended)
 
-For production applications requiring maximum accuracy, we recommend using our Python package which implements the exact preprocessing used during training:
+For maximum accuracy, use our Python package which implements the exact preprocessing used during training:
 
 ```bash
 pip install fashn-human-parser
@@ -76,10 +79,13 @@ pip install fashn-human-parser
 ```python
 from fashn_human_parser import FashnHumanParser
 
-parser = FashnHumanParser(device="cuda")
-segmentation = parser.predict(image)
+parser = FashnHumanParser()  # auto-detects GPU
+segmentation = parser.predict("image.jpg")
+# segmentation is a numpy array of shape (H, W) with class IDs 0-17
 ```
 
+The package uses `cv2.INTER_AREA` for resizing (matching training), while the HuggingFace pipeline uses PIL LANCZOS.
+
 ## Label Definitions
 
 | ID | Label |