Research map / Image generation / Autoregressive and latent generation
Object-Centric Slot Diffusion
Jindong Jiang, Fei Deng, Gautam Singh, Sungjin Ahn
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 |
|---|---|---|---|
| use diffusion decoder (LSD) instead of transformer decoder (SLATE+) | slot-based object-centric model · MOVi-E | mBO: 22.17 -> 38.96 | Improved |
| use diffusion decoder (LSD) instead of SLATE | slot-based object-centric model · CLEVRTex | mIoU: 49.54 -> 62.52 | Improved |
| use diffusion decoder (LSD) instead of SLATE+ | slot-based object-centric model · CLEVRTex | FG-ARI: 70.71 -> 64.41 | Hurt |
| use diffusion decoder (LSD) instead of SLATE+ | slot-based object-centric model · MOVi-C | FG-ARI: 52.04 -> 51.98 | No clear effect |
| LSD instead of SLATE for property prediction | slot representations · CLEVRTex | material accuracy: 69.73 -> 75.56 | Improved |
| LSD instead of SLATE+ for 3D bounding box | slot representations · MOVi-C | 3D B-Box MSE: 1.44 -> 1.44 | No clear effect |
| LSD instead of SLATE for category prediction | slot representations · MOVi-E | category accuracy: 38.93 -> 42.96 | Improved |
| LSD instead of SLATE+ for compositional generation | slot-based generator · FFHQ | FID: 98.76 -> 27.83 | Improved |
| LSD instead of SLATE+ for compositional generation | slot-based generator · MOVi-E | FID: 126.51 -> 64.76 | Improved |
| apply LSD to simple CLEVR images instead of SLATE+ | LSD · CLEVR | mIoU: 66.62 -> 37.49 | Hurt |
| train LSD on CLEVR plus CLEVRTex (Mix) (ablation) | LSD · CLEVR | mIoU: 37.49 -> 59.76 | Improved |
| train LSD on CLEVR plus CLEVRTex (Mix) (ablation) | LSD · CLEVR | FID: 16.22 -> 19.45 | Hurt |
Closest papers by meaning
- Slot-Guided Adaptation of Pre-trained Diffusion Models for Object-Centric Learning and Compositional Generation (ICLR 2025 · similarity 0.93)
- GLASS: Guided Latent Slot Diffusion for Object-Centric Learning (CVPR 2025 · similarity 0.92)
- SPOT: Self-Training with Patch-Order Permutation for Object-Centric Learning with Autoregressive Transformers (CVPR 2024 · similarity 0.88)
- LaVin-DiT: Large Vision Diffusion Transformer (CVPR 2025 · similarity 0.88)
- Dual Diffusion for Unified Image Generation and Understanding (CVPR 2025 · similarity 0.87)
- PartCrafter: Structured 3D Mesh Generation via Compositional Latent Diffusion Transformers (NeurIPS 2025 · similarity 0.87)
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.
