An Image compression simulator that uses Source Extractor and Monte Carlo methods to examine the post compressive effects different compression algorithms have.

Last update: Dec 11, 2021

Overview

ImageCompressionSimulation

An Image compression simulator that uses Source Extractor and Monte Carlo methods to examine the post compressive effects of different compression algorithms.

Author's Note

This work done behind this program was to investigate the compression effects of different compression algorithms on an real science image which then has over 10000 simulated copies (due to the lack of real data) of it via Monte Carlo methods. These simulated images are then analyzed by Source Extractor- https://sextractor.readthedocs.io/. to examine the pre and post compressive effects on the image.

To be quite honest, this program will probably be outdated and non-functioning by the time someone gets around ACTUALLY trying to run it... The main purpose is to look at legacy code so that I can implement the simulation feature into my other open source project MagnaPY- www.github.com/parkji30/MagnaPY

Some Cool Results from ICS

Some simulated stars using ICS

Analysis showing the compression effects of the HCOMPRESS algorithm. We see that the distribution of the ellipticity value detected by source extractor did not get distorted from our Monte Carlo simulation of over 1000 simulated stars as shown above. We conclude that the HCOMPRESS algorithm can be safely used for sources of this nature in our science data.

Signal to noise plot showing the limitations of Source Extractor. Essentially Source Extractor cannot detect sources with a FWHM less than 1.5 sigma.

An Image compression simulator that uses Source Extractor and Monte Carlo methods to examine the post compressive effects different compression algorithms have.

Related tags

Overview

ImageCompressionSimulation

Author's Note

Some Cool Results from ICS

Owner

James Park

EvoJAX is a scalable, general purpose, hardware-accelerated neuroevolution toolkit

Semi-supervised Semantic Segmentation with Directional Context-aware Consistency (CVPR 2021)

PyTorch implementation of "Supervised Contrastive Learning" (and SimCLR incidentally)

Pytorch code for "State-only Imitation with Transition Dynamics Mismatch" (ICLR 2020)

Implementation of the paper: "SinGAN: Learning a Generative Model from a Single Natural Image"

a grammar based feedback fuzzer

Simple, but essential Bayesian optimization package

Unified API to facilitate usage of pre-trained "perceptor" models, a la CLIP

IAST: Instance Adaptive Self-training for Unsupervised Domain Adaptation (ECCV 2020)

PSTR: End-to-End One-Step Person Search With Transformers (CVPR2022)

Python PID Tuner - Based on a FOPDT model obtained using a Open Loop Process Reaction Curve

A Lightweight Experiment & Resource Monitoring Tool 📺

More Photos are All You Need: Semi-Supervised Learning for Fine-Grained Sketch Based Image Retrieval

In this project, two programs can help you take full agvantage of time on the model training with a remote server

Node-level Graph Regression with Deep Gaussian Process Models

Feed forward VQGAN-CLIP model, where the goal is to eliminate the need for optimizing the latent space of VQGAN for each input prompt

Automatically download the cwru data set, and then divide it into training data set and test data set

Unofficial Alias-Free GAN implementation. Based on rosinality's version with expanded training and inference options.

Fantasy Points Prediction and Dream Team Formation

This is the code related to "Sparse-to-dense Feature Matching: Intra and Inter domain Cross-modal Learning in Domain Adaptation for 3D Semantic Segmentation" (ICCV 2021).