Bobo~ comments

Results 11 comments of


Bobo~

Feature requests

HR_OCR net, thanks

I load pretrained model and met error: ValueError: Layer #142 (named "gru_forward"), weight <tf.Variable 'gru_forward_2/kernel:0' shape=(512, 192) dtype=float32_ref> has shape (512, 192), but the saved weight has shape (1024, 192).

hello, i got the same error when i just use the pretrained model to predict, and i resized the test image shape to (1024,1024), thanks

About the position of coord conv layers in the regression network?

@AIprogrammer " In order to re-duce the loss of efficiency as much as possible, we do notchange convolutional layers in backbone, and only replacethe 1x1 convolution layer in FPN and...

Using EfficientNet backbone leads to precision and recall equal to zero

@farzadips did you set nc = (your dataset class number) in configs/model_efficienet.yaml?

Using EfficientNet backbone leads to precision and recall equal to zero

@farzadips hi, your nc=1? the nc in head should same as nc in data.yaml. Some people have encountered this problem before, and it is all caused by this

Using EfficientNet backbone leads to precision and recall equal to zero

@farzadips can you share your dataset? According to the current description, I can't find the problem.

Using EfficientNet backbone leads to precision and recall equal to zero

> Yes sure, can I send it via email? If yes please share with me your email or send an email to [[email protected]](mailto:[email protected]) . hello, because your dataset is too...

Using EfficientNet backbone leads to precision and recall equal to zero

@farzadips I think you just need add pretrained=True in configs/model_efficientnet.yaml ` backbone: type: efficientnet version: b1 # b1 to b8 and l2 pretrained: True `

Can't I use another pre-trained model after changing the backbone?

of course, such as hwen you use resnet-bakcbone, you can't use yolov5 pre-trained model

Can't I use another pre-trained model after changing the backbone?

> Thank you for your answer. What I want to ask is, if I use swin Tiny as the backbone, can I use swin_tiny_patch4_window7_224.pth as the pretrained model, I tried...