Facial Landmark Detection
Facial Landmark Detection is a computer vision task that involves detecting and localizing specific points or landmarks on a face, such as the eyes, nose, mouth, and chin. The goal is to accurately identify these landmarks in images or videos of faces in real-time and use them for various applications, such as face recognition, facial expression analysis, and head pose estimation.
( Image credit: Style Aggregated Network for Facial Landmark Detection )
Papers
Showing 1–25 of 139 papers
Benchmark Results
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | 3DDFA | NME | 7.01 | — | Unverified |
| 2 | Pose-Invariant | NME | 6.3 | — | Unverified |
| 3 | CFSS | NME | 5.76 | — | Unverified |
| 4 | SAN GT | NME | 3.98 | — | Unverified |
| 5 | TS3 | NME | 3.49 | — | Unverified |
| 6 | Adaloss | NME | 3.31 | — | Unverified |
| 7 | CHR2C (Inter-ocular Norm) | NME | 3.3 | — | Unverified |
| 8 | CNN-CRF (Inter-ocular Norm) | NME | 3.3 | — | Unverified |
| 9 | DCFE (Inter-ocular Norm) | NME | 3.24 | — | Unverified |
| 10 | 3DDE (Inter-ocular Norm) | NME | 3.13 | — | Unverified |
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | HPRNet (Hourglass-104) | keypoint AP | 75.4 | — | Unverified |
| 2 | HPRNet (DLA) | keypoint AP | 74.6 | — | Unverified |
| 3 | HPRNet (Hourglass-104) | keypoint AP | 59.4 | — | Unverified |
| 4 | HPRNet (DLA) | keypoint AP | 55.2 | — | Unverified |
| 5 | HPRNet (Hourglass-104) | keypoint AP | 50.4 | — | Unverified |
| 6 | HPRNet (DLA) | keypoint AP | 47 | — | Unverified |
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | DCFE (Box height Norm, 19 landmarks - no earlobs) | Mean NME | 2.17 | — | Unverified |
| 2 | 3DDE (Box height Norm, 19 landmarks - no earlobs) | Mean NME | 2.01 | — | Unverified |
| 3 | SAN | Mean NME | 1.91 | — | Unverified |
| 4 | AnchorFace | Mean NME | 1.56 | — | Unverified |
| 5 | FiFA | Mean NME | 0.92 | — | Unverified |
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | SAN | Mean NME | 1.85 | — | Unverified |
| 2 | AnchorFace | Mean NME | 1.38 | — | Unverified |
| 3 | FiFA | Mean NME | 0.8 | — | Unverified |
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | ELD (MobileNetV2) | NME | 3.09 | — | Unverified |
| 2 | ELD (EfficientNetV2B0) | NME | 2.98 | — | Unverified |
| 3 | ELD (EfficientNetV2S) | NME | 2.83 | — | Unverified |
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | ELD (EfficientNetV2B1) | NME | 4.65 | — | Unverified |
| 2 | FiFA | NME (inter-ocular) | 3.89 | — | Unverified |
| 3 | D-ViT | NME | 3.75 | — | Unverified |
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | CPM+SBR+PAM | AUC0.08 private | 59.39 | — | Unverified |
| 2 | CPM+SBR | AUC0.08 private | 58.22 | — | Unverified |
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | AnchorFace | Mean NME | 3.72 | — | Unverified |
| 2 | TS3 | Mean NME | 3.49 | — | Unverified |
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | JVCR | GTE | 7.28 | — | Unverified |