Self Dataset

[Alamnoor]

Can we use for our own dataset? Please comment if possible. The research community will be thankful to have your step wise suggestions for new datasets. Thanks

[dmburd]

The code in this repo will work with datasets that have structure compatible with the structure of ShanghaiTech. The readme file (the main page of the repo) shows the directory structure. "*.mat" are Matlab files that contain annotations in the simple form -- the coordinates of the points corresponding to centers of people's heads (or visible parts of heads) for each image. The ShanghaiTech dataset was introduced in the paper "Single-Image Crowd Counting via Multi-Column Convolutional Neural Network", you can find the details there.
Of course, you can choose any file format for storing the annotations, not necessarily the Matlab format. The code should be slightly modified according to your choices of file formats and files / directories naming.
The ground truth density maps can be generated using "gen_density_maps.py" from this repo or using any code mentioned in the awesome list ("GT generation" section).

[Alamnoor]

The code in this repo will work with datasets that have structure compatible with the structure of ShanghaiTech. The readme file (the main page of the repo) shows the directory structure. "*.mat" are Matlab files that contain annotations in the simple form -- the coordinates of the points corresponding to centers of people's heads (or visible parts of heads) for each image. The ShanghaiTech dataset was introduced in the paper "Single-Image Crowd Counting via Multi-Column Convolutional Neural Network", you can find the details there.
Of course, you can choose any file format for storing the annotations, not necessarily the Matlab format. The code should be slightly modified according to your choices of file formats and files / directories naming.
The ground truth density maps can be generated using "gen_density_maps.py" from this repo or using any code mentioned in the awesome list ("GT generation" section).

Thank you so much for your quick response. I check it and prepare my data as you suggested. Thanks again