spherical-inr 1.0.0a1

A package for spherical positional encoding

Spherical-Implicit-Neural-Representation

Spherical-Implicit-Neural-Representation unifies Fourier features (SIREN) [1], pure Spherical Harmonics (SphericalSirenNet) [3], and our learnable Herglotz‐map encodings [2] into a single PyTorch toolbox. Build implicit neural representations on:

• $\mathbb{S}^2$ (HerglotzNet, SphericalSirenNet)
• Volumetric data in $\mathbb{R}^3$ with radial basis (solid harmonics)
• Generic ℝᵈ inputs via FourierPE, HerglotzPE

Coordinate conventions:

• Angles: $\theta \in [0,\pi], \varphi ∈[0,2\pi)$ in radians.
• Full spherical: $(r, \theta, \varphi)$.

📦 Installation

pip install spherical-inr

OR for development:

git clone https://github.com/yourusername/spherical_inr.git
cd spherical_inr
pip install -e .

🚀 Quickstart

import torch
from spherical_inr import HerglotzNet

# Create a HerglotzNet: harmonic order L → num_atoms=(L+1)**2
model = HerglotzNet(
    num_atoms = 50,                # spherical-harmonic degree
    mlp_sizes=[64,64],  # two hidden layers of width 64
    output_dim=1,       # scalar output per direction
)
# Random spherical angles (θ,φ)
x = torch.rand(16,2) * torch.tensor([torch.pi, 2*torch.pi])
y = model(x)

---

📚 References

1. Sitzmann, M., Martel, J., Berg, R., Lindell, D. B., & Wetzstein, G. (2021). Implicit Neural Representations with Periodic Activation Functions (SIREN). Advances in Neural Information Processing Systems (NeurIPS). https://arxiv.org/abs/2006.09661
2. Hanon, T., et al. (2025). Herglotz-NET: Implicit Neural Representation of Spherical Data with Harmonic Positional Encoding. arXiv preprint arXiv:2502.13777. https://arxiv.org/abs/2502.13777
3. Rußwurm, M., Klemmer, K., Rolf, E., Zbinden, R., & Tuia, D. (2024). Geographic Location Encoding with Spherical Harmonics and Sinusoidal Representation Networks. arXiv preprint arXiv:2310.06743. https://arxiv.org/abs/2310.06743