forgetest5
/
PaddleOCR
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity
PaddleOCR/doc/doc_en/benchmark_en.md

3.2 KiB
Raw Blame History

BENCHMARK

This document gives the performance of the series models for Chinese and English recognition.

TEST DATA

We collected 300 images for different real application scenarios to evaluate the overall OCR system, including contract samples, license plates, nameplates, train tickets, test sheets, forms, certificates, street view images, business cards, digital meter, etc. The following figure shows some images of the test set.

MEASUREMENT

Explanation:

  • v1.0 indicates DB+CRNN models without the strategies. v1.1 indicates the PP-OCR models with the strategies and the direction classify. slim_v1.1 indicates the PP-OCR models with prunner or quantization.

  • The long size of the input for the text detector is 960.

  • The evaluation time-consuming stage is the complete stage from image input to result output, including image pre-processing and post-processing.

  • Intel Xeon 6148 is the server-side CPU model. Intel MKL-DNN is used in the test to accelerate the CPU prediction speed.

  • Snapdragon 855 is a mobile processing platform model.

Compares the model size and F-score:

Model Name Model Size
of the
Whole System(M)		 Model Size
of the Text
Detector(M)		 Model Size
of the Direction
Classifier(M)		 Model Size
of the Text
Recognizer M		 F-score
ch_ppocr_mobile_v1.1 8.1 2.6 0.9 4.6 0.5193
ch_ppocr_server_v1.1 155.1 47.2 0.9 107 0.5414
ch_ppocr_mobile_v1.0 8.6 4.1 - 4.5 0.393
ch_ppocr_server_v1.0 203.8 98.5 - 105.3 0.4436

Compares the time-consuming on T4 GPU (ms):

Model Name Overall Text Detector Direction Classifier Text Recognizer
ch_ppocr_mobile_v1.1 137 35 24 78
ch_ppocr_server_v1.1 204 39 25 140
ch_ppocr_mobile_v1.0 117 41 - 76
ch_ppocr_server_v1.0 199 52 - 147

Compares the time-consuming on CPU (ms):

Model Name Overall Text Detector Direction Classifier Text Recognizer
ch_ppocr_mobile_v1.1 421 164 51 206
ch_ppocr_mobile_v1.0 398 219 - 179

Compares the model size, F-score, the time-consuming on SD 855 of between the slim models and the original models:

Model Name Model Size
of the
Whole System(M)		 Model Size
of the Text
Detector(M)		 Model Size
of the Direction
Classifier(M)		 Model Size
of the Text
Recognizer M		 F-score SD 855
ms
ch_ppocr_mobile_v1.1 8.1 2.6 0.9 4.6 0.5193 306
ch_ppocr_mobile_slim_v1.1 3.5 1.4 0.5 1.6 0.521 268