Enhance WindowedDinov2 input validation and fix tensor reshaping

[developer0hye]

Description

Enhance WindowedDinov2 input validation and fix tensor reshaping. Added an assertion to ensure input dimensions are compatible with windowed attention, and corrected the reshaping of pixel tokens for proper window handling.

Type of change

Please delete options that are not relevant.

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• This change requires a documentation update

How has this change been tested, please provide a testcase or example of how you tested the change?

import io
import requests
import supervision as sv
from PIL import Image
from rfdetr import RFDETRBase

model = RFDETRBase(resolution=546)
CLASS_NAMES = model.class_names

url = "https://media.roboflow.com/notebooks/examples/dog-2.jpeg"

image = Image.open(io.BytesIO(requests.get(url).content))
detections = model.predict(image, threshold=0.5)

labels = [
    f"{CLASS_NAMES[class_id]} {confidence:.2f}"
    for class_id, confidence
    in zip(detections.class_id, detections.confidence)
]

annotated_image = image.copy()
annotated_image = sv.BoxAnnotator().annotate(annotated_image, detections)
annotated_image = sv.LabelAnnotator().annotate(annotated_image, detections, labels)

sv.plot_image(annotated_image)

Any specific deployment considerations

Docs

• Docs updated? What were the changes: