Mohamed Abdullah

I face the same problem But I reduced the threshold to 0.0001 and I see many bounding boxes. so try to reduce the threshold and see your confidence

@thtrieu I reached to loss ~1.6 with training on 32 classes But the confidence for all the objects is still 0.0 that mean that the model didn't learn anything. How...

Thanks @AliaksandrSiarohin for your reply. this is an example of my input and output.  what I mean that when the movement begins to be extreme towards any direction (up,...

yes, I need to train on my own dataset. but how can I use **vox-256** as pretrained model and retrain on top of it instead of retraining from scratch?

Hello @AliaksandrSiarohin I retrained on my dataset but unintentionally the training is interrupted. I want to resume training from the last saved checkpoint during the training. so, where can I...

I trained additional 26 epochs on my dataset. this is the log.txt file ``` 00000099) perceptual - 93.26425; equivariance_value - 0.15666; equivariance_jacobian - 0.49131 00000100) perceptual - 87.63938; equivariance_value -...

@AliaksandrSiarohin please any help would be appreciated.

> Just Dimension problem. As above comments mentioned, we just need to permute dims. _**torch 1.10.0 ---> x1,x2 shape[B,H,W,C] --->result shape [B,H,W,1]**_ self.feat_A = self.feat_A.permute(0, 2, 3, 1) self.feat_B =...

Hello @ritikjain51, I also need to retrain on my custom dataset. Did you come with any solution about that?

I think you can use data labelling application like labelme for example and it generates the labels as json files