Simulation Dataset for BBHnet

NOTE: OLD README, UPDATE IN PROGRESS

We generate simulation dataset to train BBHnet, our deep learning framework for detection of compact binary coalescene (CBC) gravitational-wave (GW) signals .

Example

To generate a noise dataset, simply run generateRealNoise.py:

python generateRealNoise.py -t0 1186729980 -t1 1186734086 -t0-psd 1186729980 -t1-psd 1186734086
    -fs 1024 -fmin 20 -o test_noise.h5

To also add CBC signals, enable the flag -S and add the prior distribution file in Bilby format with -p

python generateRealNoise.py -t0 1186729980 -t1 1186734086 -t0-psd 1186729980 -t1-psd 1186734086
    -fs 1024 -fmin 20 -S -p config/priors/nonspin_BBH.prior -o test_signal.h5

A full list of generateRealNoise.py arguments can be found below:

usage: generateRealNoise.py [-h] -t0 FRAME_START -t1 FRAME_STOP -t0-psd FRAME_START_PSD -t1-psd FRAME_STOP_PSD -o OUTFILE [-S]
                            [-fs SAMPLE_RATE] [-fmin HIGH_PASS] [-T SAMPLE_DURATION] [-dt TIME_STEP] [-p PRIOR_FILE]
                            [--correlation-shift CORRELATION_SHIFT] [--min-trigger MIN_TRIGGER] [--max-trigger MAX_TRIGGER]
                            [-s SEED]

optional arguments:
  -h, --help            show this help message and exit
  -t0 FRAME_START, --frame-start FRAME_START
                        starting GPS time of strain
  -t1 FRAME_STOP, --frame-stop FRAME_STOP
                        stopping GPS time of strain
  -t0-psd FRAME_START_PSD, --frame-start-psd FRAME_START_PSD
                        starting GPS time of strain for PSD estimation
  -t1-psd FRAME_STOP_PSD, --frame-stop-psd FRAME_STOP_PSD
                        stopping GPS time of strain for PSD estimation
  -o OUTFILE, --outfile OUTFILE
                        path to write output file in HDF5 format
  -S, --signal          Enable to add GW signal on top of background noise
  -fs SAMPLE_RATE, --sample-rate SAMPLE_RATE
                        sampling rate of strain
  -fmin HIGH_PASS, --high-pass HIGH_PASS
                        frequency of highpass filter
  -T SAMPLE_DURATION, --sample-duration SAMPLE_DURATION
                        duration in seconds of each sample
  -dt TIME_STEP, --time-step TIME_STEP
                        time step size in seconds between consecutive samples
  -p PRIOR_FILE, --prior-file PRIOR_FILE
                        path to prior config file. Required for signal simulation
  --correlation-shift CORRELATION_SHIFT
                        if given, also compute the correlation with given shift value
  --min-trigger MIN_TRIGGER
                        mininum trigger time w.r.t to sample. must be within [0, sample_duration]
  --max-trigger MAX_TRIGGER
                        maximum trigger time w.r.t to sample. must be within [0, sample_duration]
  -s SEED, --seed SEED  random seed for reproducibility

Simulation code and tutorial for BBHnet training data

Related tags

Overview

Simulation Dataset for BBHnet

NOTE: OLD README, UPDATE IN PROGRESS

Example

Owner

Metadata-Extractor - Metadata Extractor Script can be used to read in exif metadata

A Library for Modelling Probabilistic Hierarchical Graphical Models in PyTorch

Yoga - Yoga asana classifier for python

Code for 'Single Image 3D Shape Retrieval via Cross-Modal Instance and Category Contrastive Learning', ICCV 2021

CLADE - Efficient Semantic Image Synthesis via Class-Adaptive Normalization (TPAMI 2021)

Data and code for the paper "Importance of Kernel Bandwidth in Quantum Machine Learning"

DECA: Detailed Expression Capture and Animation (SIGGRAPH 2021)

Publication describing 3 ML examples at NSLS-II and interfacing into Bluesky

An implementation of the BADGE batch active learning algorithm.

This is the code for the paper "Motion-Focused Contrastive Learning of Video Representations" (ICCV'21).

[ICCV 2021] Official Pytorch implementation for Discriminative Region-based Multi-Label Zero-Shot Learning SOTA results on NUS-WIDE and OpenImages

RaftMLP: How Much Can Be Done Without Attention and with Less Spatial Locality?

A modular, research-friendly framework for high-performance and inference of sequence models at many scales

This repository contains part of the code used to make the images visible in the article "How does an AI Imagine the Universe?" published on Towards Data Science.

An automated facial recognition based attendance system (desktop application)

StyleGAN-NADA: CLIP-Guided Domain Adaptation of Image Generators

Pytorch implementation of AngularGrad: A New Optimization Technique for Angular Convergence of Convolutional Neural Networks

A system for quickly generating training data with weak supervision

Multi-task Learning of Order-Consistent Causal Graphs (NeuRIPs 2021)

PyTorch implementation of "A Full-Band and Sub-Band Fusion Model for Real-Time Single-Channel Speech Enhancement."