Handwritten Text Recognition On Iam
المقاييس
CER
WER
النتائج
نتائج أداء النماذج المختلفة على هذا المعيار القياسي
جدول المقارنة
| اسم النموذج | CER | WER |
|---|---|---|
| decoupled-attention-network-for-text | 6.4 | 19.6 |
| full-page-handwriting-recognition-via-image | 6.3 | - |
| pay-attention-to-what-you-read-non-recurrent | 7.62 | - |
| rethinking-text-line-recognition-models | 2.75 | - |
| start-follow-read-end-to-end-full-page | 6.4 | 23.2 |
| htr-vt-handwritten-text-recognition-with | 4.7 | 14.9 |
| full-page-handwriting-recognition-via-image | 6.7 | - |
| trocr-transformer-based-optical-character | 4.22 | - |
| no-padding-please-efficient-neural | 6.6 | 15.9 |
| rethinking-text-line-recognition-models | 2.96 | - |
| pay-attention-to-what-you-read-non-recurrent | 4.67 | - |
| trocr-transformer-based-optical-character | 2.89 | - |
| full-page-handwriting-recognition-via-image | 6.5 | - |
| end-to-end-handwritten-paragraph-text | 4.32 | 16.24 |
| dtrocr-decoder-only-transformer-for-optical | 2.38 | - |
| easter2-0-improving-convolutional-models-for | 6.21 | - |
| evaluating-sequence-to-sequence-models-for | 4.87 | - |
| trocr-transformer-based-optical-character | 3.42 | - |