Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2023
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2310.17075 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910484123353088 |
|---|---|
| author | Babu, Sudarshan Liu, Richard Zhou, Avery Maire, Michael Shakhnarovich, Greg Hanocka, Rana |
| author_facet | Babu, Sudarshan Liu, Richard Zhou, Avery Maire, Michael Shakhnarovich, Greg Hanocka, Rana |
| contents | We introduce HyperFields, a method for generating text-conditioned Neural Radiance Fields (NeRFs) with a single forward pass and (optionally) some fine-tuning. Key to our approach are: (i) a dynamic hypernetwork, which learns a smooth mapping from text token embeddings to the space of NeRFs; (ii) NeRF distillation training, which distills scenes encoded in individual NeRFs into one dynamic hypernetwork. These techniques enable a single network to fit over a hundred unique scenes. We further demonstrate that HyperFields learns a more general map between text and NeRFs, and consequently is capable of predicting novel in-distribution and out-of-distribution scenes -- either zero-shot or with a few finetuning steps. Finetuning HyperFields benefits from accelerated convergence thanks to the learned general map, and is capable of synthesizing novel scenes 5 to 10 times faster than existing neural optimization-based methods. Our ablation experiments show that both the dynamic architecture and NeRF distillation are critical to the expressivity of HyperFields. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2310_17075 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | HyperFields: Towards Zero-Shot Generation of NeRFs from Text Babu, Sudarshan Liu, Richard Zhou, Avery Maire, Michael Shakhnarovich, Greg Hanocka, Rana Computer Vision and Pattern Recognition We introduce HyperFields, a method for generating text-conditioned Neural Radiance Fields (NeRFs) with a single forward pass and (optionally) some fine-tuning. Key to our approach are: (i) a dynamic hypernetwork, which learns a smooth mapping from text token embeddings to the space of NeRFs; (ii) NeRF distillation training, which distills scenes encoded in individual NeRFs into one dynamic hypernetwork. These techniques enable a single network to fit over a hundred unique scenes. We further demonstrate that HyperFields learns a more general map between text and NeRFs, and consequently is capable of predicting novel in-distribution and out-of-distribution scenes -- either zero-shot or with a few finetuning steps. Finetuning HyperFields benefits from accelerated convergence thanks to the learned general map, and is capable of synthesizing novel scenes 5 to 10 times faster than existing neural optimization-based methods. Our ablation experiments show that both the dynamic architecture and NeRF distillation are critical to the expressivity of HyperFields. |
| title | HyperFields: Towards Zero-Shot Generation of NeRFs from Text |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2310.17075 |