Quickly determine the gender of a face. Trained on hundreds of thousands of IMDB images.
Possible Use Cases
- Adding gender classification to a face detection system
- Targeted messaging to your users based on their gender/interests
- Especially useful if you do not ask what a users gender is during registration
- Automated filter recommendations on image messaging apps like Instagram
- Smart tagging in photo gallery apps
- Categorizing profiles from social networks by gender
The input should contain a base64 encoded image or image url of a face. The image must be at least 64 x 64 pixels. In order to get the best results, make sure your input image is a tight face shot.
{
"image_base64": "BASE_64_ENCODED_IMAGE"
}
or
{
"image_url": "IMAGE_URL"
}
The output will map each gender to a percentage. The percentages measure how confident the model is that the person is of that gender.
{
"woman": 0.91,
"man": 0.09
}
The model was trained by the B-IT-BOTS robotics team on over 500,000 IMDB images.
See the MAINTAINERS.md to learn more about the underlying model as well as how to contribute to the app.