PV-mSME: Calorimeter Image Datasets for Parity Violation Search at the LHC

Calorimeter image data used in a paper investigating parity violation signatures at the Large Hadron Collider. The dataset contains three independent splits for training, validation, and testing, stored in HDF5 format. It was created by C. G. Lester from the University of Cambridge.

Use Cases

• Training anomaly detection models based on calorimeter image data to identify potential parity-violating events.
• Developing computer vision classifiers for particle physics based on unrolled eta-phi surface images.
• Benchmarking image-based analysis techniques for high-energy physics experiments using the provided train/test/private_test splits.

Strengths

• Data is structured into three independent splits (train, test, private_test) for robust model evaluation.
• Images represent a standardized unrolled eta-phi surface with a consistent shape of (n, 32, 32).

Limitations

• Row count and total dataset size are unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.

Provenance

Source

University of Cambridge

Collection Method

Likely generated from simulated or recorded particle collision events at the Large Hadron Collider.