348 followers
In this article, we studied the effect of different hyperparameter configurations on a #VisualAttention architecture composed of a pre-trained #ResidualNeuralNetwork encoder, and an #LSTM decoder for an image captioning task. @UniYachayTech https://t.co/z