Research map / Image generation / Autoregressive and latent generation
Kubric: A Scalable Dataset Generator
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Fine-tune U²-Net on the easy multi-view SOD dataset | U²-Net · Multi-view SOD easy | maxFβ: 0.825 -> 0.975 | Helped |
| Fine-tune U²-Net on the hard multi-view SOD dataset | U²-Net · Multi-view SOD hard | maxFβ: 0.594 -> 0.909 | Helped |
| Add Kubric-generated synthetic examples to MoveNet training | MoveNet · COCO, Active, and Yoga pose estimation | keypoint Mean-Average-Precision | Helped |
| Train the contrastive point tracker instead of using the no-motion, no-occlusion baseline | Point tracking · MOVi-E | error rate at the largest tracking threshold: 25.2% -> 5.8% | Helped |
| Evaluate the point tracker on vertically flipped videos | Contrastive point tracker · MOVi-E vertically flipped videos | tracking performance: about 4% | Hurt |
| Train LASR with Kubric ground-truth optical flow rather than estimated flow | LASR · Video-based 3D reconstruction | reconstruction performance | Helped |
Closest papers by meaning
- The Unmet Promise of Synthetic Training Images: Using Retrieved Real Images Performs Better (NeurIPS 2024 · similarity 0.84)
- Improving the Scaling Laws of Synthetic Data with Deliberate Practice (ICML 2025 · similarity 0.83)
- EvolveDirector: Approaching Advanced Text-to-Image Generation with Large Vision-Language Models (NeurIPS 2024 · similarity 0.83)
- Real-Fake: Effective Training Data Synthesis Through Distribution Matching (ICLR 2024 · similarity 0.83)
- Navigating Dataset Documentations in AI: A Large-Scale Analysis of Dataset Cards on HuggingFace (ICLR 2024 · similarity 0.83)
- Efficient Dataset Distillation for Pre-Trained Self-Supervised Models via Statistical Flow Matching (NeurIPS 2026 · similarity 0.83)
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.