Boosting Self-Supervision for Single-View Scene Completion via Knowledge Distillation
원문 초록
Inferring scene geometry from images via Structure from Motion is a long-standing and fundamental problem in computer vision. While classical approaches and, more recently, depth map predictions only focus on the visible parts of a scene, the task of scene completion aims to reason about geometry even in occluded regions. With the popularity of neural radiance fields (NeRFs), implicit representations also became popular for scene completion by predicting so-called density fields. Unlike explicit approaches. e.g. voxel-based methods, density fields also allow for accurate depth prediction and novel-view synthesis via image-based rendering. In this work, we propose to fuse the scene reconstruction from multiple images and distill this knowledge into a more accurate single-view scene reconstruction. To this end, we propose Multi-View Behind the Scenes (MVBTS) to fuse density fields from multiple posed images, trained fully self-supervised only from image data. Using knowledge distillation, we use MVBTS to train a single-view scene completion network via direct supervision called KDBTS. It achieves state-of-the-art performance on occupancy prediction, especially in occluded regions.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
MonoMVSNet: Monocular Priors Guided Multi-View Stereo Network
Always Clear Depth: Robust Monocular Depth Estimation under Adverse Weather
Cameras as Relative Positional Encoding
Beyond Appearance: Geometric Cues for Robust Video Instance Segmentation
LighthouseGS: Indoor Structure-aware 3D Gaussian Splatting for Panorama-Style Mobile Captures
RoboScape: Physics-informed Embodied World Model
토론
이 논문과 연결된 포럼 글이 아직 없습니다.