Diverse and accurate image description using a variational auto-encoder with an additive Gaussian encoding space

Keyphrases

• Image Description 100%
• Variational Autoencoder 100%
• Conditional Variational Autoencoder 100%
• Gaussian Prior 75%
• Model Use 25%
• Latent Space 25%
• Gaussian Mixture Model 25%
• Vanilla 25%
• K-components 25%
• Image Content 25%
• Novel Additives 25%
• Model Priors 25%
• Image Caption Generation 25%

Engineering

• Gaussians 100%
• Autoencoder 80%
• Gaussian Mixture Model 20%
• Image Content 20%

Computer Science

• Autoencoder 100%
• Gaussian Mixture Model 25%
• Image Content 25%
• Long Short-Term Memory Network 25%