https://github.com/farrajota/caltech-extract-data-toolbox

Caltech pedestrian dataset data extraction code.
https://github.com/farrajota/caltech-extract-data-toolbox

Last synced: about 12 hours ago
JSON representation

Caltech pedestrian dataset data extraction code.

• Host: GitHub
• URL: https://github.com/farrajota/caltech-extract-data-toolbox
• Owner: farrajota
• License: bsd-3-clause
• Created: 2016-05-27T10:59:35.000Z (over 8 years ago)
• Default Branch: master
• Last Pushed: 2016-11-03T11:14:55.000Z (about 8 years ago)
• Last Synced: 2023-10-20T03:58:11.915Z (about 1 year ago)
• Language: Matlab
• Size: 1.34 MB
• Stars: 8
• Watchers: 1
• Forks: 5
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Caltech's functions for data extraction

Extracts all images and annotations of the (Caltech/INRIA/Tud-Brussels/ETH/Daimler) Pedestrian dataset from data files with the `.vbb` format into `.jpg` and `.txt` files, respectively.

## Requirements

- Matlab (R2012a or above)

## Usage 

To use this code simply call **extract_file(`data_folder_path`, `dataset_name`, `store_folder_name`)** to extract the desired dataset.

Parameters default values:

- `data_folder_path = ../../`;

- `dataset_name=usa`;

- `store_folder_name=/extracted_data`

Current available datasets for extraction:

- Caltech Pedestrian

- INRIA Pedestrian

- ETH Pedestrian

- TUD-Brussels Pedestrian

- Daimler Pedestrian

## Disclamer

This code has been adapted from pdollar's [toolbox](https://github.com/pdollar/toolbox) and [evaluation code](http://www.vision.caltech.edu/Image_Datasets/CaltechPedestrians/code/code3.2.1.zip).