All Questions
16 questions
1vote
0answers
68views
Training a straight "copy the input to the output" autoencoder for audio is strangely slow
As a learning exercise, I'm training a "perfect" audio autoencoder. It has a hidden layer just as wide as the input layer, with linear activation. The expectation is that the network should ...
1vote
1answer
491views
Conditional variational autoencoder: Feeding labeled MNIST to encoder with Keras
I am looking for a code implementation of a CVAE using MNIST in Keras. I found this Youtube video: https://youtu.be/8wrLjnQ7EWQ that does VAE, but I am not sure how do I convert this and make encoder ...
2votes
1answer
840views
Improve Convolutional Autoencoder
I just built a Convolutional Autoencoder to try to reconstruct a time series with shape (4000, 10, 30). This is the code, I used a batch size of 32, but I think it ...
- 53
1vote
0answers
177views
Is there any problem with the following Python+TF+Keras code for a custom loss function and network?
I am trying to code a custom loss function for variational autoencoder. I am not using mse for reconstruction loss since I am not learning p(x|z) ~ N(mu,I). Instead ...
- 1,101
1vote
0answers
118views
Sampling for the encoder part of the VAE
my question regards the code utilized to implement the sampling function in the encoder part of VAE. Supposing that we chose a latent dimension of 2. Before the latent representation, we have 4 ...
3votes
1answer
2kviews
KL divergence loss goes to zero while training VAE
I'm trying to train a variational autoencoder to perform unsupervised classification of astronomical images (they are of size 63x63 pixels). I'm using an encoder with 2 convolutional layers and a ...
2votes
0answers
272views
Keras: DepthwiseConv3DTranspose or doing transposed Conv. with a Conv. layer
I am building an autoencoder for 3D images and would like to use Depthwise convolutions. For the encoder, I found an implementation of a depthwise 3D convolutional layer (DepthwiseConv3D). For the ...
- 432
6votes
0answers
153views
Unable to transform (greatly performing) Autoencoder into Variational Autoencoder
Following the procedure described in this SO question, I am trying to transform my (greatly performing) convolutional Autoencoder into a Variational version of the same Autoencoder. As explained in ...
2votes
0answers
149views
Results are too good.. what is wrong? How to predict correctly?
I am about to evaluate a neural network and want to check whether the predictions make sense. The variables: ...
- 570
2votes
1answer
1kviews
Keras - Autoencoder different from Encoder + Decoder
I build a CNN 1d Autoencoder in Keras, following the advice in this SO question, where Encoder and Decoder are separated. My goal is to re-use the decoder, once the Autoencoder has been trained. The ...
- 21
1vote
1answer
959views
Autoencoder or layer-based dimensionality reduction?
I have a few TB of wide data. I want to reduce the number of features in my dataset before feeding my dataset into a classification model... or should I not? Obviously, I will want to try both ...
- 549
1vote
1answer
3kviews
Autoencoder gets ~0% accuracy / doesn't train at all
So I wanted to get into the topic of 'Autoencoder', and just tested how well it would work on random vectors of size 200. ...
- 63
1vote
1answer
2kviews
How to set the Reconstruction error threshold for anomaly detection using autoencoders?
Hi I am doing anomaly detection using auto encoders.I have trained the model using 'Non Anomalous' values.Now when I give anomalous points as test data. What should be the Reconstruction error ...
- 141
-1votes
1answer
230views
Denoising Autoencoder Parameter Search
I have ran a hyperparameter search for a denoising autoencoder and the results suggest I should make the sizes of my hidden layers as large as possible (within the range of values I allowed it to ...
- 179
1vote
1answer
345views
More weightage to a categorical feature for an Autoencoder model
I am using autoencoder for anomaly detection. I don't have any labels already and so its unsupervised. If I have categorical variables, I usually one hot encode before giving it to the model. I would ...
- 245
