Skip to content

Using transfer learning on pretrained image models to learn density map generation and count the number of people in an image.

Notifications You must be signed in to change notification settings

stanleykywu/covid-19-crowd-counting

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

COVID-19-Crowd-Counting

Link to project report: https://docs.google.com/document/d/1urggk9yMcrpUDIhJftpAdjG-CM6nM6o1q43KzL9jgng/edit?usp=sharing

Requirements

To install and make sure you have all requirements necessary to run our code, you can run pip install -r requirements.txt from the root directory to verify necessary packages and install missing ones.

Training Models

There are 4 scripts to train our models:

  • Baseline Classification: train_baseline_classification.py
  • VGG16 Classification Transfer Learning: train_vgg16_classification_script.py
  • ResNet50 Density Map Generation: train_resnet50_script.py
  • VGG16 Density Map Generation: train_vgg16_script.py

Evaluating Models:

Models can be evaluated on their saved losses using evaluate_classification_training.py and evaluate_regression_training.py. Models can also be evaluated on training/validation/testing data with evaluate_classification_model.py --model='model_name' and evaluate_classification_regression_model.py --model='model_name'

VGG16 Density Map Generation Demo

Alt text

About

Using transfer learning on pretrained image models to learn density map generation and count the number of people in an image.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages