variational autoencoder
A generative model that uses variational inference to learn compact representations of data, aiming to generate new data points by sampling from the learned latent space, while also capturing uncertainty in the data distribution.
- Adaptively Coordinating with Novel Partners via Learned Latent Strategies
- BEAST: Efficient Tokenization of B-Splines Encoded Action Sequences for Imitation Learning
- Boosting Generative Image Modeling via Joint Image-Feature Synthesis
- Direct3D-S2: Gigascale 3D Generation Made Easy with Spatial Sparse Attention
- Efficient Rectified Flow for Image Fusion
- IPSI: Enhancing Structural Inference with Automatically Learned Structural Priors
- Limited Preference Data? Learning Better Reward Model with Latent Space Synthesis
- LoRA-EnVar: Parameter-Efficient Hybrid Ensemble Variational Assimilation for Weather Forecasting
- Manipulating 3D Molecules in a Fixed-Dimensional E(3)-Equivariant Latent Space
- PocketSR: The Super-Resolution Expert in Your Pocket Mobiles
- Sparc3D: Sparse Representation and Construction for High-Resolution 3D Shapes Modeling
- StableGuard: Towards Unified Copyright Protection and Tamper Localization in Latent Diffusion Models
- TARFVAE: Efficient One-Step Generative Time Series Forecasting via TARFLOW based VAE