Sample RankingRecallPrecisionHmean
Sensetime V175.00%85.71%80.00%
Sogou_MM75.00%75.00%75.00%
gts-swint-cocoonly-s4-3_175.00%75.00%75.00%
SRC-B-MachineLearningLab62.50%83.33%71.43%
SRC-B-MachineLearningLab62.50%71.43%66.67%
Tencent Youtu62.50%71.43%66.67%
Baidu IDL62.50%62.50%62.50%
Baidu VIS62.50%62.50%62.50%
Baidu VIS v262.50%62.50%62.50%
CPN (multi-scale)62.50%62.50%62.50%
FOTS50.00%80.00%61.54%
Baidu VIS v2+50.00%66.67%57.14%
Yunos_Robot1.050.00%57.14%53.33%
HoText_v150.00%57.14%53.33%
Alibaba-PAI50.00%57.14%53.33%
e2e-mlt + bug fix50.00%44.44%47.06%
TextProposals + DictNet37.50%60.00%46.15%
Deep2Text-MO37.50%42.86%40.00%
e2e-mlt37.50%42.86%40.00%
Baidu IDL v337.50%37.50%37.50%
Smile37.50%37.50%37.50%
SRC-B-TextProcessingLab25.00%33.33%28.57%
Beam search CUNI +S (decoding for TextSpotter - no postprocessing)25.00%28.57%26.67%
M_FOTS_Orig25.00%28.57%26.67%
Beam search CUNI (decoding for TextSpotter - spell checked)12.50%100.00%22.22%
Char-level PTSD on SynthText12.50%14.29%13.33%
Baseline-TextSpotter0.00%0.00%0.00%
Baseline OpenCV3.0 + Tesseract0.00%0.00%0.00%
Megvii-Image++0.00%0.00%0.00%
NVTextSpotter0.00%0.00%0.00%