IRIS: Implicit Rendering Matters for Pose-Free Novel View Synthesis
cs.CV
Submitted: 2026-09-16
Updated: 2026-09-16
Project page: https://leo-frank.github.io/IRIS
Terminology
Sources
- LRM: Large Reconstruction Model for Single Image to 3D
- No Pose at All: Self-Supervised Pose-Free 3D Gaussian Splatting from Sparse Views
- RayZer: A Self-supervised Large View Synthesis Model
- Grounding Image Matching in 3D with MASt3R
- Muskie: Multi-view Masked Image Modeling for 3D Vision Pre-training
- Scene Representation Transformer: Geometry-Free Novel View Synthesis Through Set-Latent Scene Representations
- DINOv3
- NeRF--: Neural Radiance Fields Without Known Camera Parameters
- RGBD Objects in the Wild: Scaling Real-World 3D Object Learning from RGB-D Videos
- MVSplat: Efficient 3D Gaussian Splatting from Sparse Multi-View Images
- LEAP: Liberate Sparse-view 3D Modeling from Camera Poses
- LVSM: A Large View Synthesis Model with Minimal 3D Inductive Bias
- No Pose, No Problem: Surprisingly Simple 3D Gaussian Splats from Sparse Unposed Images
- Stereo Magnification: Learning View Synthesis using Multiplane Images
Related papers
- Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
- AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
- Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
- MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
- TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
- A Survey on Efficient Vision-Language-Action Models