Disentangled Sequential Autoencoder

Abstract

We present a VAE architecture for encoding and generating high dimensional sequential data, such as video or audio. Our deep generative model learns a latent representation of the data which is split into a static and dynamic part, allowing us to approximately disentangle latent time-dependent features (dynamics) from features which are preserved over time (content). This architecture gives us partial control over generating content and dynamics by conditioning on either one of these sets of features. In our experiments on artificially generated cartoon video clips and voice recordings, we show that we can convert the content of a given sequence into another one by such content swapping. For audio, this allows us to convert a male speaker into a female speaker and vice versa, while for video we can separately manipulate shapes and dynamics. Furthermore, we give empirical evidence for the hypothesis that stochastic RNNs as latent state models are more efficient at compressing and genera

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keyli2018disentangled

Related papers

• A Multimodal Dynamical Variational Autoencoder For Audiovisual Speech Representation Learning (2023)2.26
• Semi-recurrent Cnn-based VAE-GAN For Sequential Data Generation (2018)10.21
• Switching Variational Auto-encoders For Noise-agnostic Audio-visual Speech Enhancement (2021)7.16
• RAVE: A Variational Autoencoder For Fast And High-quality Neural Audio Synthesis (2021)0.00
• Towards Robust Unsupervised Disentanglement Of Sequential Data -- A Case Study Using Music Audio (2022)0.00
• Learning And Controlling The Source-filter Representation Of Speech With A Variational Autoencoder (2022)7.50
• Learning Latent Representations For Speech Generation And Transformation (2017)13.50
• Audio-visual Speech Enhancement Using Conditional Variational Auto-encoders (2019)13.65