Is it best to apply a bounding box to the largest, unobstructed view (which would be a similar approach to labeling objects at the edge of a picture where the complete object is not within the image) or is it best to apply a bounding box to the entire object (which could include some obstructions).
Is it best to apply a bounding box to the largest, unobstructed view (which would be a similar approach to labeling objects at the edge of a picture where the complete object is not within the image) or is it best to apply a bounding box to the entire object (which could include some obstructions).