Research map / 3D vision / 3D generation
CARFF: Conditional Auto-encoded Radiance Field for 3D Scene Forecasting
Jiezhi Yang, Khushi P Desai, Charles Packer, Harshil bhatia, Nicholas Rhinehart, Rowan McAllister, Joseph E Gonzalez
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 |
|---|---|---|---|
| sample n=10 latents from the mixture density network in the controller (versus 2 or 35) (ablation) | CARFF controller · Multi-Scene Intersection planning | success rate (30 trials): 17/30 and 19/30 -> 30/30 on the hard case | Improved |
| use probabilistic CARFF controller instead of overconfident or underconfident controller | CARFF controller · Multi-Scene Two Lane Merge planning | success rate (30 trials): 0/30 -> 30/30 | Improved |
| add convolutional layer to the ViT-based PC-VAE encoder (ablation) | PC-VAE · Multi-Scene Two Lane Merge | SVM accuracy (%): 29.67 -> 88.33 | Improved |
| add convolutional layer to the ViT-based PC-VAE encoder (ablation) | PC-VAE · Multi-Scene Approaching Intersection | SVM accuracy (%): 83.83 -> 89.17 | Improved |
| freeze pretrained ViT encoder (ablation) | PC-VAE · Multi-Scene Approaching Intersection | SVM accuracy (%): 29.83 -> 89.17 | Improved |
| use pretrained ViT encoder instead of MobileNet (ablation) | PC-VAE · Multi-Scene Approaching Intersection | train PSNR: 19.37 -> 26.47 | Improved |
| add pose conditioning to the VAE decoder (ablation) | VAE · Multi-Scene Approaching Intersection | novel-view PSNR: 9.84 -> 26.37 | Improved |
| use CARFF's PC-VAE instead of NeRF-VAE | NeRF-VAE · CARLA datasets | training stability | Improved |
| increase KL weight with delayed linear schedule (ablation) | PC-VAE · CARLA datasets | latent quality | Improved |
Closest papers by meaning
- Semi-Supervised Vision-Centric 3D Occupancy World Model for Autonomous Driving (ICLR 2025 · similarity 0.86)
- Multi-Level Neural Scene Graphs for Dynamic Urban Environments (CVPR 2024 · similarity 0.86)
- COME: Adding Scene-Centric Forecasting Control to Occupancy World Model (NeurIPS 2025 · similarity 0.86)
- GaussianWorld: Gaussian World Model for Streaming 3D Occupancy Prediction (CVPR 2025 · similarity 0.86)
- OpenHype: Hyperbolic Embeddings for Hierarchical Open-Vocabulary Radiance Fields (NeurIPS 2025 · similarity 0.86)
- Visual Point Cloud Forecasting enables Scalable Autonomous Driving (CVPR 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.
