-
Notifications
You must be signed in to change notification settings - Fork 45.8k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
what the field "keep_aspect_ratio_resizer" means in the .config file? #1794
Comments
Hi @wxp0329 - Specifying the keep_aspect_ratio_resizer follows the image resizing scheme described in the Faster R-CNN paper. In this case it always resizes an image so that the smaller edge is 600 pixels and if the longer edge is greater than 1024 edges, it resizes such that the longer edge is 1024 pixels. The resulting image always has the same aspect ratio as the input image. |
@jch1 I understand now , I admire you very much . |
closing as the question was addressed. |
So would it be okay for images to be bigger than this size? |
Does it have anything to do with the annotations? i.e, do the annotations get disturbed because of this? because my dataset has images of 1080x1920 resolution and resizing it to 600 and 1024 is okay? sorry if the question is silly. this is a newbie. |
Hi sxr3455 , did you figure this out? I have the same query and was not able to find anything. |
+1 |
@jch1 Hi,I want to make sure if my input image is 640 x 352 ,then after image_resizer will the input image still be 640 x 352? |
@amin07 The bounding box coordinates are normalised to the range [0 .. 1], so resizing the images won't affect those annotations. EDIT: |
@tasercake Thanks |
hi, just to clarify with min=600 and max=800 the CNN input size is then 800x800 square, right? |
I just wanna know if I use keep_aspect_ratio_resizer in .config file , it means I will resize the input images to same aspect ratio but not a fixed shape like 224*224 , right? |
model {
faster_rcnn {
num_classes: 37
image_resizer {
keep_aspect_ratio_resizer {
min_dimension: 600
max_dimension: 1024
}
}
feature_extractor {
type: 'faster_rcnn_resnet101'
first_stage_features_stride: 16
}
on the above,what the min_dimension or max_dimension represents ? Please interprete it ,thank you very much.
The text was updated successfully, but these errors were encountered: