There are many face image datasets for image recognition, but I'm looking for one that focuses on images that show the mouth, and are tagged with if the face is showing their teeth or not. Could be smiling or making a face like Jack Nicholson in The Shining.

I'd also take a selfie image dataset, and then I could do the teeth-tagging myself.

enter image description here

enter image description here

enter image description here


2 Answers 2


One potential dataset is the Selfie Data Set

Selfie dataset contains 46,836 selfie images annotated with 36 different attributes divided into several categories as follows.

Gender: is female.

Age: baby, child, teenager, youth, middle age, senior.

Race: white, black, asian.

Face shape: oval, round, heart.

Facial gestures: smiling, frowning, mouth open, tongue out, duck face.

Hair color: black, blond, brown, red. Hair shape: curly, straight, braid.

Accessories: glasses, sunglasses, lipstick, hat, earphone.

Misc.: showing cellphone, using mirror, having braces, partial face. Lighting condition: harsh, dim.

There are a few tags that may be useful, for example, see the bold ones above.

Reference: Mahdi M. Kalayeh, Misrak Seifu, Wesna LaLanne, Mubarak Shah, How to Take a Good Selfie?, in Proceedings of ACM Multimedia Conference 2015 (ACMMM 2015), Brisbane, Australia, October 26-30, 2015.

enter image description here


Have you ever found a good teeth dataset? I'm looking for the same!

I found this project: https://github.com/zllrunning/face-parsing.PyTorch

Which can create segmentation for the open mouth, but not for the teeth.

I know iOS AVCapturePhoto API class has the best teeth segmentation mask I ever saw, but as far as I know, it only works in iOS devices.

So in theory, you can create a teeth dataset by taking portrait photos using iOS camera app in Portrait mode, which come embedded with a few segmentation masks, including teeth.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.