Jian Ye
Jian Ye
@zcswdt 官网最近无法下载,且发邮件也到不了他们信箱上,是否能给我重新分享一下百度云链接呢?谢谢
@zcswdt 你好,十分感谢,但是我这边显示分享取消了,不知道你能否传到google driver上? > > @zcswdt 官网最近无法下载,且发邮件也到不了他们信箱上,是否能给我重新分享一下百度云链接呢?谢谢 > > 链接:https://pan.baidu.com/s/112llrK1k_CS8NzvaFyXNvQ > 提取码:3kop 我没有下载下来,在我百度网盘里,你给网盘账号发一下,我分享给你吧、
@zcswdt 我的百度账号是 15071402611,邮箱[email protected] 感谢。 @seekingdeep 我无法下载,需要可用的ICDAR2017-MLT下载链接
@Therock90421 put your detection results of characters as pseudo labels, and refer to the json format of COCO.
@haythemBD hey, our model only aims to detect texts on a scene images.
@haythemBD yes
@zihao-lu 有的 和mask rcnn里的检测是一样的
@congjianting 字符的帮助很大,是主要的提升点。
@sahbak if you want to detect other languages, pls fine-tune our model on corresponding datasets.
@wulitaotao1 1. the detection code is same for all datasets, and you only need to modify some configs. 2. different datasets have different format of gt. To generate gt for...