Q-GAIN Python package enables ML and physics-informed analysis for cold-atom experiments

The Q-GAIN (quantum gas analysis and inference) Python package provides tools for deploying machine learning and physics-informed analysis techniques in cold-atom experiments. It supports classification, object detection, and feature detection in images of atomic Bose-Einstein condensates.

• Implements a module-based workflow covering data loading, preprocessing, ML feature identification, and conventional analysis.
• Demonstrated on MNIST handwritten digit classification to validate the basic framework.
• Re-implements the SolDet package for detecting solitonic excitations in time-of-flight data.
• Includes an object-detection tool for identifying quantized vortices in ring-shaped BEC images.