Welcome to SDM_YOLO!
Making the visual information of urban available to all
The Dataset
We provide a set of images of urban scenarios with more than 50000 labels Free to use
Home of urban dataset
This is a project of the "Secretaría distrital de movilidad de Bogotá, Colombia"
Download
The dataset is available here:
https://www.dropbox.com/sh/5h52fldg5yjjs1c/AAC8_AGw7x7F3AVyUr0ksdWia?dl=0
Code
In this github page you can find all the source code we develop based in the Yolo project. https://ieeexplore.ieee.org/abstract/document/8730234
Publication
If you use this project in any academic or commertial research it will be nice if you cite us in :
*
A. Forero and F. Calderon, "Vehicle and pedestrian video-tracking with classification based on deep convolutional neural networks," 2019 XXII Symposium on Image, Signal Processing and Artificial Vision (STSIVA), Bucaramanga, Colombia, 2019, pp. 1-5. doi: 10.1109/STSIVA.2019.8730234 keywords: {convolutional neural nets;image classification;image sequences;object detection;object tracking;pedestrians;road vehicles;traffic engineering computing;video signal processing;video sequences;classification algorithm;tracking algorithm;pedestrian video-tracking;deep convolutional neural networks;vehicle video-tracking;You Only Look Once method;Training;Clustering algorithms;Signal processing algorithms;Indexes;Object detection;Convolutional neural networks;Taxonomy;image processing;video object tracking;video-tracking;Object detection;vehicle counting.},
This is a forked version of the darknet repo,
To use it: The datasets to train this are available in www.urban-dataset.com relinked to this site ( https://www.dropbox.com/sh/5h52fldg5yjjs1c/AAC8_AGw7x7F3AVyUr0ksdWia?dl=0) Video tutorial: https://www.youtube.com/watch?v=E05a3SzVjho
Forked from:
Darknet
Darknet is an open source neural network framework written in C and CUDA. It is fast, easy to install, and supports CPU and GPU computation.
For more information see the Darknet project website.
For questions or issues please use the Google Group.