Car-Studio: Learning Car Radiance Fields from Single-View and Endless In-the-wild Images

TL;DR

Car-Studio introduces a novel pipeline for learning detailed, editable car radiance fields from single-view, in-the-wild images, enabling realistic and controllable urban scene rendering for autonomous driving simulations.

Contribution

The paper presents a new method for learning vehicle radiance fields from unconstrained images, addressing blurring issues and enabling sharp, editable car models in urban scene simulations.

Findings

Achieves competitive performance with baseline models.

Enables controllable appearance editing of vehicles.

Provides a new dataset for in-the-wild car images.

Abstract

Compositional neural scene graph studies have shown that radiance fields can be an efficient tool in an editable autonomous driving simulator. However, previous studies learned within a sequence of autonomous driving datasets, resulting in unsatisfactory blurring when rotating the car in the simulator. In this letter, we propose a pipeline for learning unconstrained images and building a dataset from processed images. To meet the requirements of the simulator, which demands that the vehicle maintain clarity when the perspective changes and that the contour remains sharp from the background to avoid artifacts when editing, we design a radiation field of the vehicle, a crucial part of the urban scene foreground. Through experiments, we demonstrate that our model achieves competitive performance compared to baselines. Using the datasets built from in-the-wild images, our method gradually presents a controllable appearance editing function. We will release the dataset and code on https://lty2226262.github.io/car-studio/ to facilitate further research in the field.