Methodology
We transparently document the technical principles and data-building process behind Hogamdo's face analysis results
Last updated:
🧰 Core Technology Stack
Hogamdo relies on the following proven open-source and official AI models:
-
Google MediaPipe Face LandmarkerAn open-source face landmark detection framework released by Google. It extracts 478 3D facial points in real time. 공식 문서 →
-
ArcFace (DeepFace 라이브러리)A deep learning model that converts a face into a 512-dimensional vector (embedding) to measure similarity. It is one of the industry-standard face recognition models in wide use. DeepFace GitHub →
-
In-house Scoring AlgorithmFacial metrics and embedding reference data produce comparison scores and country rankings. Country coverage can vary with the analysis path and reference dataset.
⚙️ Analysis Pipeline
Web face analysis sends a JPEG version of your photo, facial landmarks and calculated metrics to the server. Some server analysis paths use temporary image files and delete them after processing. Country rankings, returned metrics, gender, selected country and language are saved to provide shareable results. The original photo is not saved with the shared result.
From the landmark coordinates we compute 11 facial-ratio metrics: face aspect ratio, eye-size ratio, inter-eye distance, nose ratio, lip thickness, jawline angle, cheekbone width, facial symmetry, and more.
The server extracts an embedding from the photo to search reference faces. If search is disabled or fails, it falls back to the existing metric and embedding scoring path.
Reference-face search results or metric comparison scores are grouped by country, depending on the analysis path. Scores are not probabilities of being liked and do not replace preference surveys.
📐 Facial Metrics Glossary
Definitions of the six core metrics used in the analysis results and statistics pages. All values are ratios computed from landmark coordinates and represent relative proportions within the face, not absolute sizes.
The ratio of face height to width. Higher values mean a longer face, lower values a rounder one.
The area of the eyes relative to the whole face. Higher values mean the eyes occupy a larger share of the face.
How sharply defined and angular the jawline contour is. Higher values mean a more defined outline.
How symmetric the left and right facial landmarks are, measured from 0 to 1. Closer to 1 means more symmetric.
Nose size relative to the face, derived from nose length and width combined.
Lip thickness relative to the face. Higher values mean fuller lips.
🌐 Data Coverage
These counts describe the public facial-metric aggregate file. Collected images and measured samples are different counts. Embedding search datasets can have different country coverage.
📊 Data Sources
The public statistics and report use the following data and interpretation limits:
- Facial metric aggregates based on public figure images grouped by country
- Country averages with separate collected-image (total_images) and measured-sample (sample_count) counts
- Country averages are neither preference surveys nor representative population samples
The public data provides country averages and sample counts. Without individual measurements and variances, these aggregates cannot establish statistical significance between ranks or accuracy in predicting popularity.
⚠️ Limitations and Caveats
To help you interpret Hogamdo's analysis results correctly, we state the following limitations:
- Affinity scores are comparisons with reference data, not validated evidence that people in a country would like the user.
- Standards of beauty are subjective and change with era and culture. This service is provided as entertainment reference material, not a scientific verdict.
- Shooting conditions such as lighting, angle, expression, and resolution can affect the analysis results.
- It must not be used as grounds to interpret any ethnicity, country, or appearance as superior or inferior.
📜 Algorithm Changelog
지역별 편향 튜닝 완료 — 8개 지역 전체에서 이상치 편차 15.2pp까지 축소
ArcFace 임베딩 도입으로 기하학 메트릭 단독 방식 대비 정확도 향상
MediaPipe Face Landmarker 기반 초기 모델 출시 (139개국)