Sample RankingRecallPrecisionHmean
HoText_v185.71%85.71%85.71%
SRC-B-MachineLearningLab85.71%85.71%85.71%
Sogou_MM85.71%85.71%85.71%
e2e-mlt + bug fix85.71%85.71%85.71%
NVTextSpotter85.71%85.71%85.71%
TextProposals + DictNet71.43%100.00%83.33%
SRC-B-TextProcessingLab71.43%100.00%83.33%
FOTS71.43%100.00%83.33%
SRC-B-MachineLearningLab85.71%75.00%80.00%
gts-swint-cocoonly-s4-3_185.71%75.00%80.00%
CPN (multi-scale)85.71%75.00%80.00%
Yunos_Robot1.071.43%83.33%76.92%
Baidu IDL v371.43%83.33%76.92%
Sensetime V171.43%83.33%76.92%
Tencent Youtu71.43%83.33%76.92%
Baidu VIS v2+71.43%83.33%76.92%
Alibaba-PAI71.43%83.33%76.92%
Baidu VIS71.43%71.43%71.43%
Baidu VIS v271.43%71.43%71.43%
Smile71.43%71.43%71.43%
e2e-mlt71.43%71.43%71.43%
Baidu IDL71.43%62.50%66.67%
M_FOTS_Orig71.43%62.50%66.67%
Megvii-Image++57.14%57.14%57.14%
Beam search CUNI +S (decoding for TextSpotter - no postprocessing)28.57%66.67%40.00%
Deep2Text-MO14.29%25.00%18.18%
Char-level PTSD on SynthText14.29%20.00%16.67%
Baseline-TextSpotter0.00%0.00%0.00%
Beam search CUNI (decoding for TextSpotter - spell checked)0.00%0.00%0.00%
Baseline OpenCV3.0 + Tesseract0.00%0.00%0.00%