PhantomKidding

icon company University of Wisconsin Milwaukee icon location Milwaukee

Results 3 comments of PhantomKidding

Wrong predictions

@markkdev How did you resolve it? I encountered the same. Firstly, I use vgg16 for global and get similar results regardless of images. Then I switch to ROI for landmark...

All predictions are similar regardless images

@veralauee I used the pretrained model, VGG-16 Global Pooling, downloaded from the model zoo. I also tried to switch to RoI using the pretrained Landmark Pooling checkpoints but encountered same...

the reference for the attention mechanism in LSTM_Attn

I believe it is similar to BahdanauAttention in https://www.tensorflow.org/tutorials/text/nmt_with_attention