Face Analysis Methods: PSL, Golden Ratio and Our Six Dimensions
Different tools answer different questions. A measured ratio describes a construction in a photo. A community tier describes a convention. A six-dimension report organizes supported observations. None becomes a universal attractiveness verdict simply by displaying a number.
PSL and True Tier retain their scales
Our PSL and True Tier tools keep their existing 0–8 scales and tier definitions. They are community-reference products. The scale is not recalibrated into a predicted human-panel mean, and a photo estimate is not a population rank.
Golden ratio retains its mathematical reference
The golden-ratio tool compares its disclosed image ratios with phi, approximately 1.618. That mathematical target is well defined. The separate claim that closer agreement always predicts greater attractiveness is not established by computing the ratio.
Symmetry and outline retain their own constructions
The symmetry tool compares paired landmarks around its fitted axis. The outline classifier compares visible lengths and widths. Their scores describe those procedures. A head turn, occlusion or lens perspective can affect the inputs, so the drawings and limitations matter.
General face analysis uses our six dimensions
Home analysis, Face Rating, AI Attractiveness, qualitative Hunter Eyes review and portrait comparisons use feature relationships, contours, visible surface, framing, expression and capture. Each paid finding requires evidence, a status, a limitation and a comparison to try. In Method v1.1, AI rates each assessable dimension from 0 to 100 and the server calculates their rounded equal-weight mean. Free on-device scans provide geometry and a weighted geometry reference score; complete six-dimension AI scoring requires the optional report. Two-person comparisons do not merge people into one total.
Choose the question first
Use a specialist tool when you want to inspect a specific angle, ratio or outline. Use the broader report when you want to understand how several visible factors work together and select practical comparisons. Use the research library to distinguish current observations from future human-rating prediction.
Inspect the evidence
Research context
These readings explain the framework and its limits; they do not validate a particular haircut or frame choice.
AI vs Human Attractiveness Ratings: Ranking Is Not Calibration
Why agreement in ranking does not mean an AI’s absolute attractiveness score matches human ratings, and how to read uncalibrated photo judgments.
Peer-reviewed · International Journal of Multimedia Information RetrievalH-ARN: Why a Beauty Rating Needs More Than an Average
Understand H-ARN distribution learning, human disagreement, model error and how Get Face Report plans to evaluate rating distributions.
See the complete assessment standard and validation status.