在ios设备中加载TF lite格式模型以进行自定义对象检测时出错

时间:2019-08-20 16:02:03

标签: ios object tensorflow detection

我使用基于MobileNetV2的转移学习训练了自定义对象检测。我将模型转换为tflite格式,以便可以在IOS设备上运行它。但是,当我在xcode中运行模型时,出现错误。

2019-08-20 16:11:07.225613+0100 ObjectDetection[5692:1119136] 
Initialized TensorFlow Lite runtime.
2019-08-20 16:11:07.280287+0100 ObjectDetection[5692:1119136] [MC] 
System group container for 
systemgroup.com.apple.configurationprofiles path is 
/var/containers/Shared/SystemGroup/systemgroup.com.apple.configu    
rationprofiles
2019-08-20 16:11:07.282075+0100 ObjectDetection[5692:1119136] [MC] 
Reading from public effective user settings.

**Failed to invoke the interpreter with error: Provided data count 
1080000 must match the required count 602112.
Failed to invoke the interpreter with error: Provided data count 
1080000  must match the required count 602112.**

0 个答案:

没有答案