SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2407.13584 · 2024

Connecting Consistency Distillation to Score Distillation for Text-to-3D Generation

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 18 functions out of this paper's own repositories and ran 17 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
LMozart/ECCV2024-GCS-BEG canonical 17 of 18
FunctionStatusWhere it lives
interpolate_fn Ran LMozart/ECCV2024-GCS-BEG/guidance/solver.py
code served (permissive licence) · get_code("6a35b62fbc80f70a")
PILtoTorch Ran LMozart/ECCV2024-GCS-BEG/utils/general_utils.py
code served (permissive licence) · get_code("95abae1a2b93399d")
batch_get_perpendicular_component Ran LMozart/ECCV2024-GCS-BEG/guidance/perpneg_utils.py
code served (permissive licence) · get_code("fb9454cc0151a3f9")
ddim_add_noise Ran LMozart/ECCV2024-GCS-BEG/guidance/sd_step.py
code served (permissive licence) · get_code("182ef37772d9a6e2")
extract_into_tensor Ran LMozart/ECCV2024-GCS-BEG/guidance/solver.py
code served (permissive licence) · get_code("a15cfd7932844ef9")
gaussian Ran LMozart/ECCV2024-GCS-BEG/utils/loss_utils.py
code served (permissive licence) · get_code("c56b7ef16f309a45")
get_perpendicular_component Ran LMozart/ECCV2024-GCS-BEG/guidance/perpneg_utils.py
code served (permissive licence) · get_code("8b06a657ede08dde")
get_rays_torch Ran LMozart/ECCV2024-GCS-BEG/scene/cameras.py
code served (permissive licence) · get_code("53a289e51949870a")
inverse_sigmoid Ran LMozart/ECCV2024-GCS-BEG/utils/general_utils.py
code served (permissive licence) · get_code("b488da571728b636")
inverse_sigmoid_np Ran LMozart/ECCV2024-GCS-BEG/utils/general_utils.py
code served (permissive licence) · get_code("700436eebf1e003c")
l1_loss Ran LMozart/ECCV2024-GCS-BEG/utils/loss_utils.py
code served (permissive licence) · get_code("ac0e42d6fbcfbbe6")
l2_loss Ran LMozart/ECCV2024-GCS-BEG/utils/loss_utils.py
code served (permissive licence) · get_code("8c3b0f873ba11813")
predicted_origin Ran LMozart/ECCV2024-GCS-BEG/guidance/solver.py
code served (permissive licence) · get_code("6df07632ebc649de")
randn_tensor Ran LMozart/ECCV2024-GCS-BEG/guidance/sd_step.py
code served (permissive licence) · get_code("1f22666d0ffc07f3")
rgb2sat Ran LMozart/ECCV2024-GCS-BEG/guidance/gcs_utils.py
code served (permissive licence) · get_code("dd19422c076362d7")
safe_normalize Ran LMozart/ECCV2024-GCS-BEG/scene/dataset_readers.py
code served (permissive licence) · get_code("1c9cf033d080b2e0")
weighted_perpendicular_aggregator Ran LMozart/ECCV2024-GCS-BEG/guidance/perpneg_utils.py
code served (permissive licence) · get_code("b216ea040465b27a")
ddim_step Not yet run LMozart/ECCV2024-GCS-BEG/guidance/sd_step.py
code served (permissive licence) · get_code("72dee17b2e482ad4")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Although recent advancements in text-to-3D generation have significantly improved generation quality, issues like limited level of detail and low fidelity still persist, which requires further improvement. To understand the essence of those issues, we thoroughly analyze current score distillation methods by connecting theories of consistency distillation to score distillation. Based on the insights acquired through analysis, we propose an optimization framework, Guided Consistency Sampling (GCS), integrated with 3D Gaussian Splatting (3DGS) to alleviate those issues. Additionally, we have observed the persistent oversaturation in the rendered views of generated 3D assets. From experiments, we find that it is caused by unwanted accumulated brightness in 3DGS during optimization. To mitigate this issue, we introduce a Brightness-Equalized Generation (BEG) scheme in 3DGS rendering. Experimental results demonstrate that our approach generates 3D assets with more details and higher fidelity than state-of-the-art methods. The codes are released at https://github.com/LMozart/ECCV2024-GCS-BEG.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2407.13584")
get_code_for_paper("2407.13584")
have("2407.13584")

Connect an agent — have() is free.