Scene Text Detection On Icdar 2015
评估指标
F-Measure
FPS
Precision
Recall
评测结果
各个模型在此基准测试上的表现结果
比较表格
| 模型名称 | F-Measure | FPS | Precision | Recall |
|---|---|---|---|---|
| real-time-scene-text-detection-with-1 | 83.1 | 44 | 90.1 | 77.2 |
| east-an-efficient-and-accurate-scene-text | 82.9 | - | 84 | 81.9 |
| wordsup-exploiting-word-annotations-for | 77 | - | 80 | 73 |
| convolutional-character-networks | 91.55 | - | 92.65 | 90.47 |
| fots-fast-oriented-text-spotting-with-a | 87.99 | - | 91 | 85.17 |
| textfusenet-scene-text-detection-with-richer | 92.23 | - | 93.96 | 90.56 |
| convolutional-character-networks | 89.66 | - | 88.88 | 90.45 |
| east-an-efficient-and-accurate-scene-text | 78.2 | - | 83.6 | 73.5 |
| textboxes-a-single-shot-oriented-scene-text | 82.9 | - | 87.8 | 78.5 |
| fast-searching-for-a-faster-arbitrarily | 87.1 | 15.7 | 89.7 | 84.6 |
| single-shot-text-detector-with-regional | 80.7 | - | 83.3 | 78.3 |
| fast-searching-for-a-faster-arbitrarily | 84.7 | 42.7 | 88.0 | 81.7 |
| convolutional-character-networks | 90.06 | - | 91.43 | 88.74 |
| fast-searching-for-a-faster-arbitrarily | 81.7 | 60.9 | 86 | 77.9 |
| 模型 15 | - | - | - | - |
| fots-fast-oriented-text-spotting-with-a | 89.84 | - | 91.85 | 87.92 |
| a-single-shot-arbitrarily-shaped-text | 86.91 | - | 86.72 | 87.09 |
| shape-robust-text-detection-with-progressive | 87.1 | - | 88.7 | 85.5 |
| real-time-scene-text-detection-with | 87.3 | - | 91.8 | 83.2 |
| pgnet-real-time-arbitrarily-shaped-text | - | - | - | - |
| detecting-oriented-text-in-natural-images-by | 78.2 | - | 79.3 | 77.0 |
| pgnet-real-time-arbitrarily-shaped-text | 53.6 | - | 70.8 | 43.0 |
| 191209629 | 90.1 | - | 92.1 | 88.2 |
| fast-searching-for-a-faster-arbitrarily | 86.3 | 31.8 | 89.2 | 83.6 |
| real-time-scene-text-detection-with-1 | 87.3 | 10 | 90.9 | 83.9 |
| convolutional-character-networks | 90.16 | - | 90.9 | 89.44 |
| textsnake-a-flexible-representation-for | 83.7 | - | 85.5 | 82 |
| fused-text-segmentation-networks-for-multi | 84.1 | - | 88.6 | 80 |
| convolutional-character-networks | 90.97 | - | 89.99 | 91.98 |
| pyramid-mask-text-detector | 89.33 | - | 91.3 | 87.43 |
| multi-oriented-scene-text-detection-via | 84.3 | - | 89.5 | 79.7 |
| shape-robust-text-detection-with-progressive-1 | 85.7 | - | 86.9 | 84.5 |
| convolutional-character-networks | 89.7 | - | 91.15 | 88.3 |
| fast-searching-for-a-faster-arbitrarily | 82.9 | 53.9 | 86.3 | 79.8 |
| scene-text-detection-with-supervised-pyramid | 87.2 | - | 88.7 | 85.8 |
| mask-r-cnn-with-pyramid-attention-network-for | 85.9 | - | 90.8 | 81.5 |
| pixellink-detecting-scene-text-via-instance | 84.5 | - | 85.5 | 83.6 |
| character-region-awareness-for-text-detection | 86.9 | - | 89.8 | 84.3 |
| mask-textspotter-an-end-to-end-trainable | 86 | - | 91.6 | 81 |
| detecting-multi-oriented-text-with-corner | 84.5 | - | 88.7 | 80.7 |
| multi-oriented-text-detection-with-fully | 75 | - | 73.1 | 76.8 |
| efficient-and-accurate-arbitrary-shaped-text | 82.6 | - | 84.9 | 80.4 |
| geometry-normalization-networks-for-accurate | 88.52 | - | 90.41 | 86.71 |