Detection benchmark
This benchmark is closed.
The Eurocity Persons Dataset detection challenge no longer accepts submissions and no new results are evaluated. The leaderboard below is a fixed snapshot of the final standings from the previous Eurocity Persons Dataset website, kept for reference. The dataset itself remains available for download.
Results are ranked by log-average miss rate (LAMR) on the reasonable subset; lower is better. Day and Night are evaluated separately.
Day
| # | Method | Submitted by | LAMR reasonable | LAMR small | LAMR occluded | LAMR all | External data | Publication | Code | Submitted on |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | SPNet w cascade | Huawei Noah AI Th... | 0.042 | 0.095 | 0.216 | 0.139 | ImageNet | link | link | 2020-03-18 |
| 2 | LSFM | Abdul Hannan Khan | 0.044 | 0.099 | 0.230 | 0.150 | ImageNet, TJU-DHD... | link | link | 2022-10-16 |
| 3 | Pedestron | IIAI, UAE | 0.051 | 0.112 | 0.254 | 0.162 | ImageNet | link | link | 2020-03-09 |
| 4 | APD | Anonymous | 0.053 | 0.124 | 0.268 | 0.173 | ImageNet | link | — | 2020-05-08 |
| 5 | SPNet w FPN | Huawei Noah AI Th... | 0.055 | 0.121 | 0.246 | 0.165 | ImageNet | link | link | 2019-10-15 |
| 6 | Pedestrian2 | Hongsong Wang | 0.056 | 0.126 | 0.266 | 0.171 | ImageNet | — | will be released as soon as the paper is accepted | 2019-11-06 |
| 7 | DAGN | DSLab | 0.059 | 0.142 | 0.263 | 0.175 | ImageNet | link | — | 2021-07-01 |
| 8 | Real-time Pedestr... | Irtiza and LiJinp... | 0.066 | 0.136 | 0.313 | 0.193 | ImageNet | link | link | 2020-01-13 |
| 9 | Irtiza and LiJinp... | Irtiza Hasan | 0.086 | 0.168 | 0.379 | 0.230 | ImageNet | link | will be made soon | 2019-12-04 |
| 10 | YOLOv3 | ECP Team | 0.097 | 0.186 | 0.401 | 0.242 | ImageNet | link | — | 2019-04-01 |
| 11 | Faster R-CNN | ECP Team | 0.101 | 0.196 | 0.381 | 0.251 | ImageNet | link | — | 2019-04-01 |
| 12 | F2DNet | Abdul Hannan Khan | 0.107 | 0.175 | 0.387 | 0.261 | ImageNet | link | link | 2021-12-29 |
| 13 | Pedestron | Jannes Scholz | 0.121 | 0.215 | 0.524 | 0.285 | — | — | — | 2024-01-05 |
| 14 | SSD | ECP Team | 0.131 | 0.235 | 0.460 | 0.296 | ImageNet | link | — | 2019-04-02 |
| 15 | Torchvision Faste... | Attila Lengyel | 0.141 | 0.296 | 0.439 | 0.309 | ImageNet | link | link | 2020-04-21 |
| 16 | R-FCN (with OHEM) | ECP Team | 0.163 | 0.245 | 0.507 | 0.330 | ImageNet | link | — | 2019-04-01 |
| 17 | YOLOv3_640 | HUI_Tsinghua-Daim... | 0.273 | 0.564 | 0.623 | 0.456 | — | — | — | 2019-05-17 |
| 18 | YOLOv3-spp | Surromind . | 0.425 | 0.679 | 0.755 | 0.586 | ImageNet | — | — | 2019-11-13 |
| 19 | YOLOv3 | Surromind . | 0.699 | 0.916 | 0.877 | 0.789 | ImageNet | — | — | 2019-11-05 |
Night
| # | Method | Submitted by | LAMR reasonable | LAMR small | LAMR occluded | LAMR all | External data | Publication | Code | Submitted on |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | SPNet w cascade | Huawei Noah AI Th... | 0.066 | 0.119 | 0.231 | 0.131 | ImageNet | link | link | 2020-03-18 |
| 2 | Pedestrian2 | Hongsong Wang | 0.071 | 0.127 | 0.244 | 0.140 | ImageNet | — | will be released as soon as the paper is accepted | 2019-11-06 |
| 3 | SPNet w FPN | Huawei Noah AI Th... | 0.090 | 0.172 | 0.292 | 0.170 | ImageNet | link | link | 2019-10-15 |
| 4 | Pedestron (retrai... | Anonymous | 0.096 | 0.158 | 0.275 | 0.162 | ImageNet, Wider P... | — | — | 2021-09-06 |
| 5 | FasterRCNN with M... | Anonymous | 0.150 | 0.253 | 0.653 | 0.295 | ImageNet | — | — | 2019-07-08 |
| 6 | Faster R-CNN | ECP Team | 0.201 | 0.359 | 0.701 | 0.358 | ImageNet | link | — | 2019-05-02 |
Some values in this snapshot are truncated.