Skip to content
/ g2net Public
generated from AnswerDotAI/nbdev-template

CNN for continuous gravitational wave detection in LIGO data

License

Notifications You must be signed in to change notification settings

DrHB/g2net

Repository files navigation

Continuous Gravitational Wave Detection

Deep Learning for LIGO Signal Classification

LIGO IOP arXiv Kaggle

CNN-based detection of continuous gravitational waves from spinning neutron stars in LIGO data


Problem

Continuous gravitational waves (CWs) are persistent signals emitted by rapidly rotating neutron stars. Unlike transient signals from black hole mergers, CWs are much weaker and require computationally expensive matched-filter searches across large parameter spaces.

This solution uses deep learning on Short Fourier Transform (SFT) spectrograms from LIGO's H1 and L1 detectors, achieving 3rd place in the G2NET competition. The winning approaches from this competition demonstrated potential to reduce computing costs by 1-3 orders of magnitude. This work contributed to the published paper "Learning to detect continuous gravitational waves".

Approach

Data Pipeline

SFT (Short Fourier Transform) files from LIGO contain complex Fourier coefficients across 360 time steps and 4096 frequency bins for each detector (H1, L1). The preprocessing:

  1. Load complex SFT data from HDF5 files
  2. Compute power spectrum: P = Re(z)² + Im(z)²
  3. Normalize by mean power per detector
  4. Compress frequency dimension: 4096 → 128 bins via averaging
  5. Stack H1/L1 as 2-channel image: (2, 360, 128)

Model

  • Architecture: ConvNeXt (medium/large variants)
  • Input: 2-channel time-frequency spectrograms
  • Output: Binary classification (signal present / absent)
  • Training: Mixed precision, cosine annealing, mixup augmentation
  • Debiasing: Post-processing to correct for dataset distribution shifts

Experiments

Folder Focus
EXP_10 Baseline CNN experiments
EXP_20 EfficientNet, PSD features
EXP_40 Noise injection, denoising
EXP_50 ConvNeXt variants (final models)
EXP_120 ViT baselines
EXP_200 Cached dataset experiments
DEBIAS_SUB Submission debiasing

Links

Keywords

gravitational-waves ligo deep-learning signal-processing cnn pytorch neutron-stars astrophysics kaggle-competition


Star this repo if you find it useful!

About

CNN for continuous gravitational wave detection in LIGO data

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published