PRM

🚀 PRM

PRM is a novel photometric stereo based large reconstruction model that can reconstruct high - quality meshes with fine - grained local details.

🚀 Quick Start

For the code implementation of this project, you can visit the GitHub repository. You can also view the pre - printed paper on Arxiv. Additionally, there is a Gradio demo available on Hugging Face.

✨ Features

We propose PRM, a novel photometric stereo based large reconstruction model to reconstruct high - quality meshes with fine - grained local details.

• Rich Photometric Cues: Unlike previous large reconstruction models that use images prepared under fixed and simple lighting as both input and supervision, PRM renders photometric stereo images by varying materials and lighting. This not only improves the precise local details but also increases the model’s robustness to variations in the appearance of input images.
• Flexible Image Rendering: To offer enhanced flexibility of images rendering, we incorporate a real - time rendering method and mesh rasterization for online images rendering.
• Differentiable PBR and Multiple Photometric Supervisions: By employing an explicit mesh as our 3D representation, PRM ensures the application of differentiable PBR, which supports the utilization of multiple photometric supervisions and better models the specular color for high - quality geometry optimization.

PRM leverages photometric stereo images to achieve high - quality reconstructions with fine - grained local details, even amidst sophisticated image appearances. Extensive experiments demonstrate that PRM significantly outperforms other models.

📄 License

This project is licensed under the Apache - 2.0 license.