Computer vision and pattern recognition research from arXiv for January 4, 2024.
Today's Themes (AI Generated)
- Segmentation methods for medical images and general objects using transformer models like SAM and CLIP.
- Image generation and editing techniques using diffusion models with text and image conditioning.
- Multi-modal transformer models for tasks like image captioning, visual QA, action recognition.
- Techniques for 3D shape and pose estimation from images or point clouds.
- Adapting large vision-language models like CLIP for downstream tasks via prompt learning.
