Hash-QNeRF: Multiresolution Hash Encoding for Quantum Neural Radiance Fields

2026-07-23

AI Summary

This work addresses the limited scalability and slow convergence of conventional Quantum Neural Radiance Fields (QNeRF) in high-complexity, high-resolution scenes, which stem from their reliance on sinusoidal positional encoding. To overcome these limitations, we propose the first integration of Instant-NGP’s multi-resolution learnable hash encoding into QNeRF, replacing the original encoding scheme. This modification substantially improves training efficiency and memory usage while preserving the quantum radiance field’s prediction pipeline and robustness to noise. Our approach combines parameterized quantum circuits, amplitude embedding, parity measurements, and volume rendering. Evaluated on the Blender synthetic dataset, it achieves a training loss of 0.003534 (corresponding to approximately 24.5 dB PSNR) and demonstrates state fidelity between 0.93 and 0.98 in Qiskit simulations, confirming its effectiveness and superiority.

Abstract

Neural Radiance Fields (NeRF) have revolutionized novel view synthesis, yet their classical implementations remain computationally intensive for high-fidelity rendering. QNeRF recently demonstrated the feasibility of training NeRF on gate-based quantum computers by combining amplitude embedding, parameterized quantum circuits (PQCs), parity-based measurements, and volumetric rendering. However, QNeRF relies on classical sinusoidal positional encoding for spatial coordinates, which scales poorly with scene complexity and resolution. In this work, we replace the sinusoidal positional encoding for spatial coordinates with the multiresolution hash encoding from Instant-NGP while keeping the view-direction encoding, amplitude MLP, quantum circuit, parity measurement, output scaling, and volumetric rendering pipeline unchanged. This hybrid design, Hash-QNeRF, retains the quantum radiance prediction step while benefiting from the fast convergence and memory efficiency of learnable hash grids. On a synthetic Blender scene, we achieve a final training loss of 0.003534, corresponding to approximately 24.5 dB PSNR on the fitted batch. Noise resilience experiments using Qiskit FakeKyiv and FakeTorino backends yield state fidelities of 0.93 to 0.98, indicating that hash encoding does not degrade the quantum circuit's noise tolerance.

Problem

• Quantum Neural Radiance Fields
• positional encoding
• scene complexity
• resolution scalability
• computational efficiency

Innovation

• multiresolution hash encoding
• Quantum Neural Radiance Fields
• parameterized quantum circuits
• noise resilience
• novel view synthesis