Skip to content

AndreaMattiaGaravagno/model_centric_track

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 

Repository files navigation

Model-Centric Track

This is the official GitHub of the Model-Centric Track of the Wake Vision Challenge (TBD: link to the challenge).

It asks participants to push the boundaries of tiny computer vision by innovating model architectures to achieve high test accuracy while minimizing resource usage, leveraging the newly released Wake Vision, a person detection dataset.

Proposed model architectures will be evaluated over a private test set.

To Get Started

Create a new environment.

python -m venv /path/to/new/virtual/environment

Activate the environment.

source /path/to/new/virtual/environment/bin/activate

Install requirements.

python -m pip install -r requirements.txt

In "model_centric_track.py" modify the value of the variable "data_dir", writing the path to the location in which you would like to save the dataset (239.25 GiB).

The first execution will require hours, since it has to download the entire dataset. It will train the ColabNAS model, a state-of-the-art person detection model, on the Wake Vision dataset to get you started. Then you can modify the "model_centric_track.py" script as you like, and propose your own model architecture.

python model_centric_track.py

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages