Replies: 3 comments 1 reply
-
最新版本中已经去掉了,更新一下就可 |
Beta Was this translation helpful? Give feedback.
0 replies
-
看到了改成这个了np.mean(conf_list if any(conf_list) else [0]))) 能不能弄一个版本,返回时增加每个字符的字符和置信度集合,自己改了改代码,可以返回结果,不知道有没有哪里写的有问题, [[[54.0, 120.0], [183.0, 121.0], [183.0, 156.0], [53.0, 155.0]], 'キレイで', 0.99378806, [('キ', 0.9998), ('レ', 0.9999), ('イ', 0.9754), ('で', 1.0)]] |
Beta Was this translation helpful? Give feedback.
0 replies
-
还有就是偶尔会发生多字符,少字符的问题,有没有什么解决办法? |
Beta Was this translation helpful? Give feedback.
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
在计算置信度平均值时,不知道为什么要加+ [1e-50],结果,我看平均值反而不对了
result_list.append((text, np.mean(conf_list + [1e-50])))
[[65.0, 122.0], [196.0, 123.0], [196.0, 158.0], [65.0, 158.0]], 'キレイで', 0.7883910775184632, [('キ', 0.9997), ('レ', 0.9995), ('イ', 0.9427), ('で', 1.0)]]
Beta Was this translation helpful? Give feedback.
All reactions