Less Decoder is More Encoder: Geometric Representation Learning from Novel View Synthesis
cs.CV, cs.AI, cs.RO
Submitted: 2026-10-02
Updated: 2026-10-02
Project page: https://snap-nvs.github.io
Terminology
Sources
- Revisiting Feature Prediction for Learning Visual Representations from Video
- Muskie: Multi-view Masked Image Modeling for 3D Vision Pre-training
- DINOv3
- Stereo Magnification: Learning View Synthesis using Multiplane Images
- MuM: Multi-View Masked Image Modeling for 3D Vision
- GeoAware-VLA: Implicit Geometry Aware Vision-Language-Action Model
- V-JEPA 2.1: Unlocking Dense Features in Video Self-Supervised Learning
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