top of page

DONT (DO Not Touch your face)


To overcome pandemic of SARS-CoV-2 in the world, there are several basic protective measures against COVID-19 to wash your hands frequently, avoid touching eyes, nose and mouth, practice respiratory hygiene, and maintain social distancing spread according to the WHO advice(

Especially, it is very important to avoid touching eyes, nose and mouth and wash your hands frequently. However, it's very difficult to wash your hand perfectly (, and people unconsciously touch their faces more than 3,000 times a day. Therefore, we want to warn the touching their face by developing an automated monitoring software with deep learning.

What it does

This program can reduce the frequency of touching your face by giving you an alarm when you touch your face unconsciously with web-camera and monitoring with deep learning technology.

How we built it

  • Firstly, The action recognition neural network (I3D) was used to recognize the video clips on the action of touching the face (ver.I3D).

  • Secondly, for light version of deep learning model, MobileNet3 was used to recognize the action of touching the face with three consequent snapshot images (ver.MobileNet).

  • It was created in the form of EXE program so that it could be easily used by ordinary users.

  • We made and publish a dataset to be used freely to develop recognition software on the action of touching the face.

Challenges we ran into

  • More data are needed for higher and robust accuracy in various kinds of situations.

  • Lighter, faster neural network model is needed to be used in smartphone and/or ordinary PC without specialized GPU, etc.

Accomplishments that we're proud of

  • The first model can recognize face-touching actions in 0.08 sec with a GTX960 or higher GPUs (ver.I3D, 92% accuracy)

  • The second model can recognize face-touching actions in 0.07 sec with Intel(R) Core i7-6700 CPU 3.40GHz or higher (ver.MobileNet, 92% accuracy)

  • We have released out pre-trained action recognition model, which can be used freely

What we learned

  • We realized that people touch their faces unconsciously.

  • Fully automated monitoring with deep learning for detecting touch-the-face could be used to prevent SARS-CoV-2 from spreading.

  • It is very important to collaborate with teammates quickly.

What's next for DONT

  • To add sound alarm.

  • To add a "report" function based on 24-hour monitoring mode.

  • To modify DONT to Apply in CCTV, CPU-only machine and mobile phones in actual environment


Built With

opencv, python, pytorch

3D Printing

Safety devices with 3D printing

  • With the pandemic of COVID 19, the health workers all over the world are scrambling with the shortage of safety devices such as masks as well as medical devices.

  • To overcome the novel corona virus pandemic, this project is aimed to share the ideas and know-hows of making newly developed devices - which not previously existed, but should be exist - as well as the ways of making up the weak points of existing devices(such as not-fitting-masks or should-contact-hand-sanitizers) which are necessary under this circumstance.

  • Uploaded STL files are produced or modified with 3D modelling program( 3-matics (Materialise), Solidworks s/w)

  • Download or Use of uploaded STL files are free and highly recommended. (Copyright 2020. Univeristy of Ulsan College of Medicine. All right reserved according to FreeBSD License.)

  • For further information or questions, please contact via email:

* This project is run by Medical Imaging and Intelligent Reality Lab in Asan Medical Center(AMC-MI2RL), Seoul, Korea.

Medical device reverse engineering

  • Various ideas to back in COVID-19 pandemic can be realized with 3D printing.

  • Unmet needs of frontline healthcare workers for medical devices can be fulfilled.

  • It can be done with following protocol,

  • get 3D image from CT or 3D scanner

  • make parts of the same shape or customized parts

  • print with appropriate 3D printing material

Related article link

1. Human-specific mask

  • This is to product personal customized mask using 3D scanning

  • Merits: cost effective, replaceable HEPA filter, sterilization available, face-fit customized(can reduce the air flow gap)

  • Printing with flexible materials (e.g. TPU in FDM) is recommended.

  • Original design of AMC-MI2RL.

  • STL link: Preparing





2. Forearm door opener

  • We propose new way of a door opening. This can reduce the chance of potential exposure occurring when touching the contaminated door or door handle surface with hand.

  • Demo video:

  • We refered modelling of STARARASYS Corp., and modified with additional modeling of MI2RL to improve user convenience and stability of installation.

  • Printing with robust materials (e.g. ABS, PLA in FDM) are recommended.

  • STL link



















3. Portable Untact Tool; PUT

  • With this small device, people can minimize the chance of potential exposure reducing the hand contact to the possibly contaminated surfaces such as door handles(when opening/closing the door) or elevator buttons.

  • How to use: Fix the index finger on the "fixed finger" part to use it. The "pressing button" is used to press the elevator button, etc., and in the "keyring holder" part, the portability has been increased so that it can be worn on name card necklace of keyring. "open/close door" part can be used when the open or close door.

  • This device is developed with AMC-MI2RL original modelling

  • Printing with robust materials (e.g. ABS, PLA in FDM) are recommended.

  • STL link:



4. Size-adjustable ear guard

  • Wearing a mask for many hours can cause pain behind the ear. This device is developed to lessen that kind of pain and also reduce the gap between face and the mask by size control.

  • Merit: Made in two type
    – Type1 : putting on around the ear, projections make resistance with hairs so that not slip, size adjustable for face size (3-step).
    – Type2: not putting on the ear, without projections so that the contact surface on the neck is not uncomfortable, adjustable for face size (3-step).

  • Printing with flexible materials (e.g. TPU in FDM) are recommended.

  • STL link:











5. Face/eye protector

  • This device is developed as a tool for protecting eyes and face from infectious virus (respiratory droplets) by covering whole face.

  • It can be used with attachment of transparent acrylic sheet (OHP film) onto the 3d printed frame.

  • 3D printed frame + Transparent Acrylic sheets OHP film.

    • OHP films can be attached(holded) onto the frame through the holes with 3-hole punch. (Spec: 93.5*93.5 mm)

  • Printing with robust materials (e.g. ABS, PLA in FDM) are recommended.

  • Original design of AMC-MI2RL.

  • Improvements: Resized to fit normal heads, flattened the surface that touches the skin so that there is no pain, and fixed with a sponge.

  • STL link:































6. Automatic hand dispenser housing

  • This device is developed as a housing unit for DON’T push dispenser.

  • It can be used with various size of hand sanitizers.

  • Original design of AMC-MI2RL.

  • Printing with robust materials (e.g. ABS, PLA in FDM, CJP) are recommended.

  • STL link:





7. Contributer

Screen Shot 2020-07-17 at 2.38.33 PM.png
Screen Shot 2020-07-17 at 2.38.43 PM.png
Screen Shot 2020-07-17 at 2.38.52 PM.png
Screen Shot 2020-07-17 at 2.39.13 PM.png
Screen Shot 2020-07-17 at 2.39.24 PM.png


COVID-19 time-dependent reproduction rate of the world​

Daily time-dependent reproduction rates are calculated per country based on the publicly released numbers of confirmed cases.



DONT push the dispenser

  • DONT push the dispenser, AVOID the chance of cross-contamination.

  • Since we understand that, as engineers, it is our responsibility to put our knowledge to overcome the difficulties of our community and society, especially in the era of global pandemic(COVID-19), we developed this "non-contact type, " "automated" pumping device for hand-sanitizer. This project is a contribution of Medical Imaging and Intelligent Reality Lab in Asan Medical Center(AMC-MI2RL) in Seoul, Korea.

  • This device is made with "Arduino Nano CH340", "ultrasonic sensor" and two pieces of "sg90 servo motors".

  • Demo Video(YouTube link)

Setting Environment and Rationale

  • Setting

    • for digital input/output TX, RX, D2~D13: 14 pin

    • for analog input A0~A7: 8 pin

    • Atmel Atmega328P-AU

    • Support USB upload and power supply as well as the external power supply(DC 5V~12V)

    • Use 5V-pin for 3.3V ~ 5.5V supply and use VIN-pin for 6V ~ 12V

      • When supplying power to the external module using 5Vpin, we recommend lower than 100 milliamperes(mA) of current

    • Warning: The diode can be burnt with the current higher than 100mA when using USB port power supply.

  • Parts needed

  • Rationale











  • When using the hand-sanitizers in public spaces, we are often exposed to a risk of touching the surface, which nobody knows who touches.

  • For whoever feared about potential exposure to contaminations through publicly used hand-sanitizers, this will be quite helpful.

  • It is recommended to rub hands with hand-sanitizers containing alcohol, especially for health workers or for the people entering the public space. The pump's surface could have been contaminated with many people and can possibly increase the risk of infection. To decrease the possibility of those risks, many uses automatic pumps. This project shares the way of making 'DIY' auto-pumping hand sanitizer.

1. NEWS(20.03.19): DONT Push Dispenser ver 0.4

In the same axis of "DONT/WASH", we developed a non-contact type of liquid dispenser to reduce the cross-contamination, which can occur when touching at the surface of pump-type hand sanitizers.


2. Circuit Construction

Connect jumper wires to the adequate port, referring to the image below.

3. Quick Start Guide


  • Setting

    1. Download the "Arduino IDE"(You can find and download it by searching from Google).

    2. Connect CH340 to PC with USB connector

    3. Connect the devices like the image above

    4. After finishing the primary setting of Arduino IDE.

    5. Open "DONT_dispenser.ino" and upload the file to Arduino with "Ctrl +u"

    6. Disconnect the USB line with the PC and connect the 9V battery to the device.

  • Run the device

    • If the hand approaches within 12cm from the ultrasound sensor, the device will automatically pump the sanitizer.

    • How to wash hand properly?


4. Project Contributors

bottom of page