why the decode result is messy

ViTAE-Transformer / DeepSolo

The official repo for [CVPR'23] "DeepSolo: Let Transformer Decoder with Explicit Points Solo for Text Spotting" & [ArXiv'23] "DeepSolo++: Let Transformer Decoder with Explicit Points Solo for Multilingual Text Spotting"

Other

242 stars 34 forks source link

why the decode result is messy #26

Closed taotaoyuhust closed 1 year ago

taotaoyuhust commented 1 year ago

the command is python demo.py --config-file ../configs/R_50/TotalText/finetune_150k_tt_mlt_13_15.yaml --input ../testpics/ --output ../testresult/ --opts MODEL.WEIGHTS ../weights/res50_pretrain_synch-art-lsvt-rects.pth