David Silva

Closing due to inactivity

Closing due to inactivity

Closing due to inactivity

No, you'll have to implement it using [predict_generator](https://keras.io/models/sequential/#predict_generator) to get the predictions. Should be similar to the [test](https://github.com/davidtvs/Keras-LinkNet/blob/master/main.py#L111) function, pass it the test data generator and it'll return the predictions...

At least for `nn.ConvTranspose2D` there's `output_padding` to address this issue, see: https://pytorch.org/docs/stable/generated/torch.nn.ConvTranspose2d.html#convtranspose2d Looks like for `nn.PixelSuffle` there's no way around it though, maybe that would be a good feature request...

You can see [here](https://github.com/davidtvs/Keras-LinkNet/blob/master/main.py#L155) how the number of classes is extracted from a batch returned by the data generator. So I would start by writing a data generator class for...

I faced the same issue on ROS 2 Humble. Even after trying different ways to specify rows and columns it still was giving me weird visualizations. I dug into the...