synthesiaresearch/humanrf
Neural radiance field system for high-fidelity 3D reconstruction of humans in motion from multi-view camera capture.
Not currently ranked — collecting fresh signals.
star history
HumanRF is a neural rendering method that reconstructs moving humans in 3D using multi-view camera sequences. It leverages neural networks to represent scene geometry and appearance as continuous radiance fields, enabling novel view synthesis of human performance. The project includes training code, data loaders for the ActorsHQ dataset, and inference pipelines for rendering free-viewpoint video of captured subjects.
Frequently asked
- What is synthesiaresearch/humanrf?
- Neural radiance field system for high-fidelity 3D reconstruction of humans in motion from multi-view camera capture.
- Is humanrf open source?
- Yes — synthesiaresearch/humanrf is an open-source project tracked on heatdrop.
- What language is humanrf written in?
- synthesiaresearch/humanrf is primarily written in Python.
- How popular is humanrf?
- synthesiaresearch/humanrf has 496 stars on GitHub.
- Where can I find humanrf?
- synthesiaresearch/humanrf is on GitHub at https://github.com/synthesiaresearch/humanrf.