Net Magnetic Moment Extraction from Noisy Magnetic Field Data using Residual Neural Networks

Quantum sensors, such as Nitrogen- Vacancy (NV) centers in diamond, allow us to image magnetic fields with high spatial resolution, making it a powerful tool in characterizing a variety of magnetic samples. However, extracting net magnetic moment information from noisy magnetic field image data is a challenging inverse problem. We address this challenge with an image-based machine learning approach. We use a Residual Neural Network (ResNet) to obtain vector net magnetic moments from noisy non-gaussian data. We train our model with synthetically generated magnetic sources superimposed on measured lab noise images. We can successfully process images with signal-to-noise ratios below 0.1. These results demonstrate a robust solution for precise characterization of magnetic sources in noisy regimes where traditional fitting techniques are insufficient.