Face shape detector

Upload one straight-on photo and this face shape detector compares your visible outline with oval, round, square, heart, oblong and diamond profiles. It draws the measured widths and shows a match for each shape, because most faces do not fit one label perfectly. The comparison stays in your browser. Use it as a styling starting point, not a judgment of facial attractiveness.

Six outline prototypes

A face-shape label summarizes how the forehead, cheekbones, jaw and chin relate to one another. It is not an anatomical diagnosis. Hair, beard growth, soft tissue and the angle of a camera alter the outline in an image. A blended result such as oval with a square lean may describe a photo better than forcing a strict single category.

Four features used by the face shape detector

The detector compares four front-view features with six hand-tuned prototypes: face length relative to cheekbone width, forehead width, jaw width and chin angle. It shows the closest shape plus match percentages for oval, round, square, heart, oblong and diamond. These percentages sum to the classifier’s own distribution; they are not probabilities from a labeled population study.

Reading close shape matches

Read the strongest match together with the runner-up. Close scores mean the photo sits between two category profiles. A high percentage means similarity to one of this tool’s preset profiles, not a probability that a person has an objectively true shape. A more useful question is whether the visible forehead, cheek and jaw widths make sense on the image.

A hypothetical face shape detector might show oval at 38% and round at 36%, with the remaining four matches sharing the rest. That near tie means two hand-tuned prototypes fit similarly under this photo; it is not a 38% chance that the person objectively has an oval face. Before using the first label to choose a haircut, compare the outline from another level photo and consider hair texture and how much forehead you prefer to show.

The prototypes emphasize different relationships. In the hand-tuned configuration, oval has a longer length-to-cheek-width target than round, while square has a wider jaw relative to cheek width. Those are examples of the classifier’s settings, not observed population averages. A face shape detector match can shift when a photo makes the forehead seem narrower or the jaw seem wider. Compare which measured guide line drove the change instead of assuming the person switched categories.

Hair and lens effects on the outline

Hair covering the outer forehead can narrow the apparent top width, while a beard can make the lower outline seem wider than the actual jaw landmarks. A close selfie also changes relative widths across depth. Pull hair away from the outline and use a level front portrait. Look at the forehead, cheek, jaw and length guide lines if the top label seems surprising. For an unobstructed outer facial outline, read how to take a photo for a PSL rating.

What the match percentages leave undecided

The percentages are relative to six selected prototypes, and their sum of 100% does not prove that any one label is anatomically exact. Hair, facial hair and soft tissue can hide the contour beneath. For a haircut choice, hair texture, growth pattern and personal preference may matter as much as the outline category.

A face shape detector result uses the same prototype settings for men and women; the percentages are not sex-based population probabilities. A reversible styling trial can compare hair lengths or partings against the two leading outline labels. The page cannot render the future haircut, so a mirror, photographs and the person’s preference remain necessary.

Using a shape shortlist for hairstyle trials

For numerical width and height context, use face proportions. The jawline rating page examines the lower outline with a separate approximation. Those measurements can help explain why a particular outline label led the list without turning it into a hair recommendation.

Face shape detector FAQ

What is my face shape if two types tie?

A blend is common. Look at the width lines and use both likely shapes as styling references.

Can a detector find a diamond face?

It compares the cheek width against narrower forehead and jaw measurements, though hair can hide those edges.

Does a beard affect the result?

Yes. Thick facial hair can make the jaw look wider or hide its contour.

Is an oval face better than a round face?

No. These names describe outlines, not attractiveness rankings.

Can I use a selfie?

Yes, but a longer camera distance reduces perspective distortion.

Is the photo uploaded?

No. This shape calculation runs entirely in your browser.