Related Experiment Video
Updated: Jun 30, 2025
09:23
Quantum State Engineering of Light with Continuous-wave Optical Parametric Oscillators
Published on: May 30, 2014
14.5K
Efficient learning of mixed-state tomography for photonic quantum walk
Qin-Qin Wang1,2, Shaojun Dong3, Xiao-Wei Li4
1CAS Key Laboratory of Quantum Information, University of Science and Technology of China, Hefei 230026, China.
Science Advances
|March 15, 2024
Summary
We developed a neural network method to reconstruct mixed states in open quantum walks (QW) using fewer measurements. This approach achieves high fidelity and improves the scalability of photonic QW experiments.
Area of Science:
- Quantum information science
- Quantum computing
- Photonic systems
Background:
- Open quantum walks (QW) leverage noise for enhanced applications.
- Verifying open QW success is difficult due to resource-intensive mixed-state tomography and physical measurement constraints.
Purpose of the Study:
- To develop a more efficient method for reconstructing mixed states in open discrete-time QW.
- To reduce the number of measurements required for state tomography.
- To improve the scalability of photonic QW setups.
Main Methods:
- Utilized a neural density operator to model the quantum system and its environment.
- Employed a generalized natural gradient descent for accelerated training.
- Introduced a compact interferometric measurement device for photonic QW.
Main Results:
- Achieved high-fidelity mixed-state reconstruction (approximately 97.5%).
- Reduced measurement requirements by 50% compared to traditional methods for 1D open discrete-time QW.
- Demonstrated experimental learning of mixed states with a scalable photonic setup.
Conclusions:
- Neural networks offer a powerful and efficient alternative to conventional state tomography.
- The proposed method enhances the practicality and scalability of open QW applications.
- This work facilitates the experimental investigation of noise-enhanced quantum phenomena.
