CODEWITHSUNDEEP
pythontensorflowtensorflow2.0object-detection
TourEiffel
TourEiffel

Reputation: 4414

Tensor Flow Zoo Model Object Detection - Does size affect result?

I am watching the list of all tensorflow2 Zoo Model. Assuming that 640x640 is the size of image, I was wondering what happen if the input image is bigger than the model size.

For example if we use :

SSD ResNet50 V1 FPN 640x640 (RetinaNet50)

If 640x640 is the image size and the image we use as input got a size of 1915x1080, would it affect quality of object detection? If yes, should we use SSD ResNet50 V1 FPN 1024x1024 (RetinaNet50) in order to get higher performances?

If my assumption is false what means 640x640 ?

Upvotes: 0

Views: 489

Answers (1)

Tim Roberts
Tim Roberts

Reputation: 54698

Ah, the age old question, "does size matter"? Yes, it does. A model that is trained on 640x640 images expects to receive only 640x640 images. If you have an image of 1915x1080, it is your job to reduce that to 640x640 for classification. Often, the biggest part of a classification app is munging varied input into the standard form the model expects.

YOU will have to decide how to shrink the image. Do you need to maintain aspect ratio? Do you leave black bars top and bottom?

And do not assume that the 1024 model necessarily does a better job than the 640 model. It is very often the case that more detail simply confuses the classification.

Upvotes: 2

Related Questions