When we buy some stuff, we are frequently encountered web user interface. for example, kiosk ordering without human cashier continue to spread, to make their workflow smarter and reduce labor costs. It means, even if we are not online. If we want to go grocery shopping, have a coffee or buy a hamburger, than have to be fluency on their user interface.
relatively young people easily dealing with it, but older people often fail to use it.
Do you think it just because of that older people less exposed to IT products? Maybe yes in some sense, but mainly because of unfriendly user interfaces for the elderly. Even their physical aspects are ignored.
Look at this lady who can not order the desired menu, because font size is too small for her.
Providing a suitable UI for each age group will contribute to improving user experience and preventing digital alienation against certain ages.
- docs/ tutorials
- webpage/ demo webpage
- static/
- templates/ html
- app.py
- create_db.py to make custom dataset form imdb-wiki
- utils.py
- build torch custom dataset
@Hong Min
- train model (on progress)
- build webpage (on progress)
FrontEnd - @Dustin Gogoll
BackEnd - @Hong Min
- publish it on hiroku (on progress)
@Akash.py
add function with live steam web cam
first detects the face in image
- Predict age from face image by Convolution Neural Networks.
- considered age estimation as classification problem
- targets are from 0 to 101 that represent age.
- Provide suitable UI according to Age group.
- e.g. contents size, way of talking, emojis ..
- IMDB-WIKI
- FGnet
- UTKFace
consists of 20k+ face images in the wild (only single face in one image)
labelled by age, gender, and ethnicity
- Adience collection of unfiltered faces
Total number of photos: 26,580
Number of age groups / labels: 8 (0-2, 4-6, 8-13, 15-20, 25-32, 38-43, 48-53, 60-)
[Deep expectation of real and apparent age from a single image without facial landmarks](https://www.vision.ee.ethz.ch/en/publications/papers/articles/eth_biwi_01299.pdf)International Journal of Computer Vision (IJCV), 2016