Splatent: A New Method for Novel View Synthesis Using Diffusion Latents

This research explores novel view synthesis using diffusion model latents, a promising area for 3D reconstruction. The paper's novelty lies in its application of 'splatting' techniques within the latent space of diffusion models.

• •Addresses the problem of novel view synthesis, a key challenge in computer vision.
• •Employs diffusion model latents, leveraging the power of generative models.
• •Introduces 'splatting' techniques within a novel latent space framework.