Deep Learning on Object-centric 3D Neural Fields
This addresses the challenge of using NFs in practical deep learning workflows for 3D object representation, though it appears incremental as it builds on existing NF methods.
The paper tackles the problem of integrating Neural Fields (NFs) into deep learning pipelines for downstream tasks by introducing nf2vec, a framework that generates compact latent representations for input NFs in a single inference pass, enabling successful application to various tasks using exclusively NFs.
In recent years, Neural Fields (NFs) have emerged as an effective tool for encoding diverse continuous signals such as images, videos, audio, and 3D shapes. When applied to 3D data, NFs offer a solution to the fragmentation and limitations associated with prevalent discrete representations. However, given that NFs are essentially neural networks, it remains unclear whether and how they can be seamlessly integrated into deep learning pipelines for solving downstream tasks. This paper addresses this research problem and introduces nf2vec, a framework capable of generating a compact latent representation for an input NF in a single inference pass. We demonstrate that nf2vec effectively embeds 3D objects represented by the input NFs and showcase how the resulting embeddings can be employed in deep learning pipelines to successfully address various tasks, all while processing exclusively NFs. We test this framework on several NFs used to represent 3D surfaces, such as unsigned/signed distance and occupancy fields. Moreover, we demonstrate the effectiveness of our approach with more complex NFs that encompass both geometry and appearance of 3D objects such as neural radiance fields.