Research map / Image generation / Fast diffusion sampling
JointDiT: Enhancing RGB-Depth Joint Modeling with Diffusion Transformers
Kwon Byung-Ki, Qi Dai, Lee Hyoseok, Chong Luo, Tae-Hyun Oh
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add adaptive scheduling weights (ablation) | JointDiT · ImageNet 6K joint generation | FID: 30.88 -> 24.20 | Improved |
| add adaptive scheduling weights (ablation) | JointDiT · MSCOCO 30K joint generation | FID: 15.17 -> 11.13 | Improved |
| add adaptive scheduling weights (ablation) | JointDiT · NYUv2 depth estimation | AbsRel: 8.8 -> 6.4 | Improved |
| add unbalanced timestep sampling (ablation) | JointDiT · NYUv2 depth estimation | AbsRel: 8.8 -> 7.8 | Improved |
| add unbalanced timestep sampling on top of adaptive weights (ablation) | JointDiT · NYUv2 depth estimation | AbsRel: 6.4 -> 5.7 | Improved |
| add unbalanced timestep sampling without adaptive weights (ablation) | JointDiT · ImageNet 6K joint generation | FID: 30.88 -> 29.37 | Improved |
| add unbalanced timestep sampling on top of adaptive weights (ablation) | JointDiT · ImageNet 6K joint generation | FID: 24.20 -> 24.26 | No clear effect |
| add unbalanced timestep sampling on top of adaptive weights (ablation) | JointDiT · ImageNet 6K joint generation | IS: 37.04 -> 37.81 | Improved |
| add adaptive scheduling weights (ablation) | JointDiT · OpenImages 6K depth-conditioned generation | FID: 11.94 -> 14.37 | Hurt |
| add unbalanced timestep sampling on top of adaptive weights (ablation) | JointDiT · OpenImages 6K depth-conditioned generation | FID: 14.37 -> 12.58 | Improved |
| add both adaptive weights and unbalanced timestep (ablation) | JointDiT · OpenImages 6K depth-conditioned generation | ImageReward Rank1: 26.35 -> 30.73 | Improved |
| increase LoRA rank from 16 to 64 (ablation) | JointDiT · NYUv2 depth estimation | AbsRel: 9.1 -> 5.7 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- ControlNet — ControlNet++: Improving Conditional Controls with Efficient Consistency Feedback (ECCV 2024) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- JointNet: Extending Text-to-Image Diffusion for Dense Distribution Modeling (ICLR 2024 · similarity 0.90)
- Pixel-Perfect Depth with Semantics-Prompted Diffusion Transformers (NeurIPS 2025 · similarity 0.87)
- A Simple Approach to Unifying Diffusion-based Conditional Generation (ICLR 2025 · similarity 0.87)
- Relay Diffusion: Unifying diffusion process across resolutions for image synthesis (ICLR 2024 · similarity 0.87)
- UDT: Reconciling U-Nets and Diffusion Transformers with Data-Adaptive Token Reduction (NeurIPS 2026 · similarity 0.86)
- Coherent 3D Scene Diffusion From a Single RGB Image (NeurIPS 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
