Advancing AI-Driven 3D Reconstruction: Implementing and Refining NeRF Technologies with Neuralangelo

Faculty Mentor

Travis Masingale

Presentation Type

Oral Presentation

Start Date

May 2025

End Date

May 2025

Location

PUB 319

Primary Discipline of Presentation

Computer Science

Abstract

Neural Radiance Fields (NeRF) have emerged as a groundbreaking technique in AI-driven 3D reconstruction, offering high-fidelity scene rendering from 2D images. This research focuses on the implementation and optimization of Neuralangelo, an advanced NeRF-based model, with an emphasis on workflow efficiency, computational accessibility, and real-world applications. A key challenge in utilizing Neuralangelo is its complex installation process and high computational demands, which can limit accessibility for researchers and developers with standard hardware. To address this, I have developed an improved dependency management system and streamlined installation process, making Neuralangelo more accessible without compromising performance. Additionally, this research explores NeRF’s potential applications in 3D asset generation, digital preservation, automotive design, and interactive media, while also examining the ethical implications of AI-driven 3D reconstruction. By refining installation workflows and evaluating Neuralangelo’s scalability, this research contributes to the broader adoption of AI-powered 3D modeling. The findings highlight the potential for NeRF-based models to revolutionize digital visualization and computational rendering, paving the way for future advancements in AI-driven 3D reconstruction.

Comments

My current research is based off of & picked up from:

@inproceedings{li2023neuralangelo, title={Neuralangelo: High-Fidelity Neural Surface Reconstruction}, author={Li, Zhaoshuo and M\"uller, Thomas and Evans, Alex and Taylor, Russell H and Unberath, Mathias and Liu, Ming-Yu and Lin, Chen-Hsuan}, booktitle={IEEE Conference on Computer Vision and Pattern Recognition ({CVPR})}, year={2023} } The link to the new release is located here at: https://github.com/beasmith152/Easy-Install-Neuralangelo-includes-updated-dependency-list-/tree/main

This document is currently not available here.

Share

COinS
 
May 7th, 10:50 AM May 7th, 11:10 AM

Advancing AI-Driven 3D Reconstruction: Implementing and Refining NeRF Technologies with Neuralangelo

PUB 319

Neural Radiance Fields (NeRF) have emerged as a groundbreaking technique in AI-driven 3D reconstruction, offering high-fidelity scene rendering from 2D images. This research focuses on the implementation and optimization of Neuralangelo, an advanced NeRF-based model, with an emphasis on workflow efficiency, computational accessibility, and real-world applications. A key challenge in utilizing Neuralangelo is its complex installation process and high computational demands, which can limit accessibility for researchers and developers with standard hardware. To address this, I have developed an improved dependency management system and streamlined installation process, making Neuralangelo more accessible without compromising performance. Additionally, this research explores NeRF’s potential applications in 3D asset generation, digital preservation, automotive design, and interactive media, while also examining the ethical implications of AI-driven 3D reconstruction. By refining installation workflows and evaluating Neuralangelo’s scalability, this research contributes to the broader adoption of AI-powered 3D modeling. The findings highlight the potential for NeRF-based models to revolutionize digital visualization and computational rendering, paving the way for future advancements in AI-driven 3D reconstruction.