Next time you go through airport security or to a doctor's appointment give feedback on the service by showing a gesture to a camera. Easy way to collect and see people's impressions of a service.
The application uses a classification model that has been trained to recongise 3 gestures (peace, devil horns and thumbs up), all the other hand gestures are recongised as "other".
NB! the dataset that was used to train the classification model is limited as it only used gestures used by the author, for a more generalised applicaiton of the application, the classificaiton model needs to be improved!!!
This was a project put together for the Hack the Middlands 7.
NB! There are probably a number of dependiencies that are not neccessary, yet the virtual envioroment with the given dependencies in the "dependencies.txt" file, should allow to launch the applicaiton.
python3 -m htm-env
python3 -m pip install -r dependencies.txt
source ./htm-env/bin/activate
- Eddit the settings,json file, change to the approprite path:
- video_ip - ip address for an ip camera.
- model - path to the model in the ./data/model folder used for classifying the gesture.
... There are exaple pathS in the settings file.
- Launch as a pytonh script and the application will be availabe on a local address. (check terminal for the address)
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent