Score Distillation via DDIM Inversion
While 2D diffusion models generate realistic, high-detail images, 3D shape generation methods like Score Distillation Sampling (SDS) built on these 2D diffusion models produce cartoon-like, over-smoothed shapes. To help explain this discrepancy, in this paper we prove that the image guidance used in...
Main Author: | Lukoianov, Artem S. |
---|---|
Other Authors: | Solomon, Justin |
Format: | Thesis |
Published: |
Massachusetts Institute of Technology
2024
|
Online Access: | https://hdl.handle.net/1721.1/156343 |
Similar Items
-
Towards Efficient Local 3D Conditioning
by: Zhang, Dingxi, et al.
Published: (2023) -
Multi-teacher Contrastive Knowledge Inversion for Data-Free Distillation
by: LIN Zhenyuan, LIN Shaohui, YAO Yiwu, HE Gaoqi, WANG Changbo, MA Lizhuang
Published: (2023-11-01) -
Produced Water Desalination via Pervaporative Distillation
by: Jingbo Wang, et al.
Published: (2020-12-01) -
Aluminothermic Reduction of Sulfides via Reactive Vacuum Distillation
by: Stinn, Caspar, et al.
Published: (2024) -
Few-Shot Image Classification via Mutual Distillation
by: Tianshu Zhang, et al.
Published: (2023-12-01)