Detecting Usage of Mobile Phones using Deep

Learning

We have used deep-learning algorithms to detect mobile phone usage. We are releasing the pretrained models and dataset here. The readme in those folders provide the necessary explanations. Also see our research paper to know more.

21-Jan-2023 update: We have developed a new dataset, named MUID-IITR, which is available here. Researchers can use this in place of IITH-GDMU dataset.

29-Nov-2022 update: The below links are not working and the authors have lost access to the dataset. If any user/researcher had earlier downloaded the dataset, we request them to share the dataset with authors again, so we can make them available to the scientific community again. Thank you so much.

Links for the datasets

IITH-GDMU dataset - Click here
KaggleDriver dataset - Click here
csv Files (provides the bounding-box labels for all the images in the dataset)- Click here
Readme file (gives an overview of the datasets and explains the format of the csv files) - Click here

Links for the pretrained models

Models pretrained on IITH-GDMU dataset - Click here
Models pretrained on KaggleDriver dataset - Click here
Readme file - Click here

Developers

This research was led by Dr Sparsh Mittal (ECE, IIT Roorkee) who was formerly at CSE, IIT Hyderabad. The students working on the research are Poonam Rajput and Subhrajit Nag (CSE, IITH) .

If you have questions, please send them to the following emails: sparshfec at iitr.ac.in, cs17mtech11019 at iith.ac.in and cs17resch11006 at iith.ac.in. We will respond to them in a best-effort manner.

If you use any of these datasets or the pretrained models, then please cite this paper:

@inproceedings {rajput2020detecting,

title = {{Detecting Usage of Mobile Phones using Deep Learning Technique}},

year = "2020",

author = "Poonam Rajput and Subhrajit Nag and Sparsh Mittal",

booktitle = "6th EAI International Conference on Smart Objects and Technologies for Social Good (GOODTECHS)",

}