Table 1 The experimental results of some state-of-the-art handcrafted descriptor-based and deep-learning-based methods on different datasets from the year 2010–2020
From: Vehicle logo detection using an IoAverage loss on dataset VLD100K-61
Author | Images | Image size | Classes | Scheme/platform | mAP | Precision | Years | Datasets |
|---|---|---|---|---|---|---|---|---|
Apostolos [13] | 1200 | 100 × 100 | 10 | SIFT + EMS | / | 94% | 2010 | / |
Chun Pan [31] | 200 | 140 × 100 | 26 | CNN | / | 99.23% | 2013 | / |
Llorca [18] | 3579 | 640 × 480 | 27 | HOG + SVM | / | 92.59% | 2013 | / |
Haoyu Peng [20] | 3370 | 50 × 50 | 16 | SRSD + MS | / | 97.21% | 2015 | / |
Yue Huang [29] | 11,500 | 70 × 70 | 10 | CNN + PS | / | 99.07% | 2015 | XMU |
Yizhang Xia [32] | 19,780 | 64 × 64 | 15 | CNN + MTL | / | 98.14% | 2016 | Xiamen University |
Ruilong Chen [22] | 11,500 | 70 × 70 | 10 | Spatial-SIFT + LR | / | 99.93% | 2016 | / |
Ruilong Chen [19] | 10,000 | 70 × 70 | 10 | CPLR | / | 98.80% | 2016 | / |
Li Huan [33] | 80,000 | 1360 × 1024 | 10 | DBNs | / | 92% | 2017 | / |
Sotheeswaran [23] | 500 | 800 × 600 | 25 | CtFS | / | 86.30% | 2018 | / |
Foo Chong [35] | 1500 | 70 × 70 | 13 | HPO-DCNN | / | 99.10% | 2018 | / |
Ruilong Chen [34] | 1500 | 70 × 70 | 10 | CN | / | 100% | 2018 | / |
Yu Ye [54] | 16,000 | 64 × 96 | 80 | OE-POEM | / | 99.10% | 2018 | HFUT-VL |
Foo Chong [36] | 11,500 | 70 × 70 | 10 | CNN + ZCA | / | 99.13% | 2019 | XMU |
Shuo Yang [37] | 2010 | 320 × 240 − 1500 × 938 | 30 | YOLOv3 | 0.899 | 2019 | VLD-30 | |
Ruikang Liu [39] | 14,950 | 10 × 10–150 × 150 | 13 | SSFPD + EMS | / | 99.52% | 2019 | CVLD |
Zhongjie [38] | 4000 | 1000 × 600 | 8 | Faster-RCNN + VGG-16 | 0.943 | / | 2019 | / |
Yongtao Yu [28] | 75,000 | / | 15 | CDCN | / | 98.10% | 2019 | / |
Junxing [41] | 3015 | 938 × 668 | 45 | Ss-VLD | 0.881 | / | 2020 | / |
Shuo Yang [42] | 2010 | 320 × 240 − 4608 × 3450 | / | Faster-RCNN | 0.875 | / | 2020 | / |
Nguyen [40] | 8000 | 720 × 1160 | 13 | Darknet-53 | 0.905 | / | 2020 | / |
Linghua Zhou [30] | 18,089 | 576 × 576 | 17 | Filter-Deblur-GAN + VL-YOLO | 0.981 | / | 2020 | LOGO-17 |
Kittikhun [25] | 4000 | / | 40 | HOG + SS | / | 75.25% | 2020 | VLR-40 |
Ye Yu [27] | 1500 | 70 × 70 | 10 | MLPNL | / | 99.98% | 2021 | |
Junxing [53] | 2010 | 320 × 240 − 1500 × 938 | 30 | F-RCNN + VLD-C | 0.911 | / | 2021 | VLD-45-S |
Shuo Yang [26] | 45,000 | 610 × 378 − 7359 × 4422 | 45 | YOLOV4 | 0.847 | / | 2021 | VLD-45 |