Calibration of Quantum Sensors by Neural Networks

Area of Science:

• Quantum technology
• Sensor development
• Machine learning applications

Background:

• Real-world quantum sensor deployment requires high reliability and controllability.
• Limited resources for calibration pose a challenge for producers and operators.
• Neural networks are widely used for classical sensor calibration.

Purpose of the Study:

• To apply neural network techniques for calibrating a quantum photonic sensor.
• To reduce calibration overhead by using only available probe states for training data.
• To investigate the impact of parameter space coverage on calibration accuracy.

Main Methods:

• Utilized a neural network approach for sensor calibration.
• Collected training data based on available probe states.
• Analyzed the effect of finely covering the parameter space.

Main Results:

• Demonstrated the successful application of neural networks to quantum photonic sensor calibration.
• Showed that comprehensive parameter space coverage is essential for minimizing uncertainties.
• Achieved calibration uncertainties close to the ultimate theoretical limits.

Conclusions:

• Neural network calibration is a viable and efficient method for quantum sensors.
• This technique can enhance the trustworthiness and usability of quantum devices in practical settings.
• The proposed method has the potential to become a standard for quantum sensor calibration.