Just letting TF reshape the input to (w, h) with the resizer, without preprocessing. Otherwise, you'll have to choose an image size (w, h) yourself more or less arbitrarily (some kind of average of your data would do). In that case, just use a fixed_shape_resizer with the same ratio. In the best case, your data always has the same width/height ratio. So you need to use a fixed shape resizer for SSD.
This is because it ends with fully connected layers, for which you need to know the size of the feature maps whereas for F-RCNN there are only convolutions (which work on any input size) up to the ROI-pooling layer (which only doesnt need a fixed image size). SSD and faster R-CNN work quite differently one from another, so, even though F-RCNN has no such constraint, for SSD you need input images that always have the same size (actually you need the feature map to always have the same size, but the best way to ensure it is with always the same input size).
0 kommentar(er)
