pals-ttic/sjc

• highreadme#1

  Reposition the README's opening to explicitly state its 3D generation purpose

  Why:

  CURRENT

  # Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation
  
  Haochen Wang*,
  Xiaodan Du*,
  Jiahao Li*,
  Raymond A. Yeh†,
  Greg Shakhnarovich
  (* indicates equal contribution)
  
  TTI-Chicago, †Purdue University
  
  Abstract: *A diffusion model learns to predict a vector field of gradients. We propose to apply chain rule on the learned gradients, and back-propagate the score of a diffusion model through the Jacobian of a differentiable renderer, which we instantiate to be a voxel radiance field. This setup aggregates 2D scores at multiple camera viewpoints into a 3D score, and repurposes a pretrained 2D model for 3D data generation. We identify a technical challenge of distribution mismatch that arises in this application, and propose a novel estimation mechanism to resolve it. We run our algorithm on several off-the-shelf diffusion image generative models, including the recently released Stable Diffusion trained on the large-scale LAION dataset.*

  COPY-PASTE FIX

  Copy

  # Score Jacobian Chaining: Lifting Pretrained 2D Diffusion Models for 3D Generation
  
  This repository provides the official implementation for Score Jacobian Chaining (SJC), a method that repurposes pretrained 2D diffusion models for high-quality 3D object generation.
  
  Haochen Wang*,
  Xiaodan Du*,
  Jiahao Li*,
  Raymond A. Yeh†,
  Greg Shakhnarovich
  (* indicates equal contribution)
  
  TTI-Chicago, †Purdue University
  
  Abstract: *A diffusion model learns to predict a vector field of gradients. We propose to apply chain rule on the learned gradients, and back-propagate the score of a diffusion model through the Jacobian of a differentiable renderer, which we instantiate to be a voxel radiance field. This setup aggregates 2D scores at multiple camera viewpoints into a 3D score, and repurposes a pretrained 2D model for 3D data generation. We identify a technical challenge of distribution mismatch that arises in this application, and propose a novel estimation mechanism to resolve it. We run our algorithm on several off-the-shelf diffusion image generative models, including the recently released Stable Diffusion trained on the large-scale LAION dataset.*

  OpenDoneDismiss
• mediumtopics#2

  Add more specific topics to align with 3D generation methods

  Why:

  CURRENT

  3d-generation, diffusion-models

  COPY-PASTE FIX

  Copy

  3d-generation, diffusion-models, text-to-3d, image-to-3d, nerf, cvpr-2023

  OpenDoneDismiss
• lowreadme#3

  Clarify the project's license in the README

  Why:

  COPY-PASTE FIX

  Copy

  ## License
  This project is licensed under the terms specified in the [LICENSE](LICENSE) file.

  OpenDoneDismiss

Category GEO backends resolved for this scan: google/gemini-2.5-flash, deepseek/deepseek-v4-flash

• CATEGORY QUERY

  How to generate 3D geometry and textures using only pretrained 2D diffusion models?

  you: not recommended
  AI recommended (in order):

  1. DreamFusion
  2. MVDream
  3. SyncDreamer
  4. Fantasia3D
  5. Imagen
  6. Stable Diffusion
  7. Zero123
  8. Zero123++
  9. COLMAP
  10. OpenMVG
  11. Shap-E
  12. Point-E
  13. Luma AI's Genie
  14. MiDaS
  15. DPT
  16. Open3D
  17. ControlNet
  18. Meshroom
  19. Metashape
  20. RealityCapture

  AI recommended 20 alternatives but never named pals-ttic/sjc. This is the gap to close.

  Show full AI answer
• CATEGORY QUERY

  Seeking methods for synthesizing 3D objects by adapting existing 2D image generation models.

  you: not recommended
  AI recommended (in order):

  1. Stable Diffusion
  2. Imagen
  3. Neural Radiance Field (NeRF)
  4. DreamFusion
  5. Magic3D
  6. Fantasia3D
  7. ProlificDreamer
  8. GET3D
  9. StyleGAN2
  10. StyleGAN3
  11. StyleGAN-NADA
  12. GIRAFFE
  13. pi-GAN
  14. CLIP
  15. CLIP-NeRF
  16. CLIP-Forge
  17. LDM3D
  18. HoloGAN
  19. VoxelGAN

  AI recommended 19 alternatives but never named pals-ttic/sjc. This is the gap to close.

  Show full AI answer

• Metadata completeness

  pass

• README presence

  pass

• Compared to common alternatives in this category, what is the core differentiator of pals-ttic/sjc?

  pass
  ⚠AI did not name pals-ttic/sjc — likely talking about a different project

  AI answers can be confidently wrong. Read for accuracy: does it match your actual tech stack, audience, and differentiator?

• If a team adopts pals-ttic/sjc in production, what risks or prerequisites should they evaluate first?

  pass
  ✓AI named pals-ttic/sjc explicitly

  AI answers can be confidently wrong. Read for accuracy: does it match your actual tech stack, audience, and differentiator?

• In one sentence, what problem does the repo pals-ttic/sjc solve, and who is the primary audience?

  pass
  ✓AI named pals-ttic/sjc explicitly

  AI answers can be confidently wrong. Read for accuracy: does it match your actual tech stack, audience, and differentiator?