MonoNeRF: Learning Generalizable NeRFs from Monocular Videos without Camera Pose

TL;DR

MonoNeRF introduces a neural radiance field model trained on monocular videos without ground-truth depth or pose, enabling depth estimation, pose estimation, and novel view synthesis from single images.

Contribution

It presents a novel Autoencoder-based architecture that learns generalizable NeRFs directly from monocular videos without supervision.

Findings

Achieves accurate depth and pose estimation from monocular videos.

Enables high-quality single-image novel view synthesis.

Operates without ground-truth annotations for training.

Abstract

We propose a generalizable neural radiance fields - MonoNeRF, that can be trained on large-scale monocular videos of moving in static scenes without any ground-truth annotations of depth and camera poses. MonoNeRF follows an Autoencoder-based architecture, where the encoder estimates the monocular depth and the camera pose, and the decoder constructs a Multiplane NeRF representation based on the depth encoder feature, and renders the input frames with the estimated camera. The learning is supervised by the reconstruction error. Once the model is learned, it can be applied to multiple applications including depth estimation, camera pose estimation, and single-image novel view synthesis. More qualitative results are available at: https://oasisyang.github.io/mononerf .