谷歌语音识别，如果选择英语（美国）、英语（爱尔兰），转录出的英文文字没有空格

例如：
1
00:00:00,000 --> 00:00:03,200
hellodragonsmynameisSamJonesI'm

2
00:00:03,200 --> 00:00:06,300
thefounderofgenerateandI'mheretodaytoask

3
00:00:06,300 --> 00:00:09,400
the£60,000inreturnfor10%

4
00:00:09,400 --> 00:00:13,100
equitywithin

5
00:00:13,100 --> 00:00:16,200
theadvertisingindustryisthatit'sbuilt

但如果只选择“英语”，不选择具体国家或地区，就没有这种问题：但准确率锐减。
例如：
1
00:00:00,000 --> 00:00:03,000
how many dragons my name is Sam Jones on the

2
00:00:03,000 --> 00:00:06,000
founder of generate and I'm here today to ask

3
00:00:06,000 --> 00:00:09,200
the 60,000 lb in return to 10%

4
00:00:09,200 --> 00:00:12,400
equity the elephant Secrets

5
00:00:12,400 --> 00:00:15,400
within the advertising industry is that is

Reproduce the transcription with Google speech recognition using English (United States), English (Ireland), and the generic English option. Compare the generated subtitle text and verify that the regional options preserve spaces between words without reducing the reported recognition accuracy.