Computer vision and pattern recognition research from arXiv for December 21, 2023.
Today's Themes (AI Generated)
- Text-to-image synthesis via diffusion models for photorealistic generation and controllable editing
- 3D reconstruction from sparse views and interactions using neural representations
- Weakly supervised learning techniques for segmentation and action localization
- Vision-language model alignment and application to embodied tasks like autonomous driving
- Generative modeling of detailed 3D head and animal shapes and motions
