TensorFlow Lite

https://github.com/tensorflow/tensorflow/tree/master/tensorflow/lite/examples/python/

TensorFlow Lite Python image classification demo

# Get photo
curl https://raw.githubusercontent.com/tensorflow/tensorflow/master/tensorflow/lite/examples/label_image/testdata/grace_hopper.bmp > /tmp/grace_hopper.bmp
# Get model
curl https://storage.googleapis.com/download.tensorflow.org/models/mobilenet_v1_2018_02_22/mobilenet_v1_1.0_224.tgz | tar xzv -C /tmp
# Get labels
curl https://storage.googleapis.com/download.tensorflow.org/models/mobilenet_v1_1.0_224_frozen.tgz  | tar xzv -C /tmp  mobilenet_v1_1.0_224/labels.txt

mv /tmp/mobilenet_v1_1.0_224/labels.txt /tmp/

首先，我们需要把该下载的图片、模型、标签，下载到本地

然后运行下面的指令:
python3 label_image.py 
  --model_file mobilenet_v1_1.0_224/mobilenet_v1_1.0_224.tflite 
  --label_file mobilenet_v1_1.0_224_frozen/mobilenet_v1_1.0_224/labels.txt 
  --image grace_hopper.bmp

输出的预测结果如下所示:
0.919720: 653:military uniform
0.017762: 907:Windsor tie
0.007507: 668:mortarboard
0.005419: 466:bulletproof vest
0.003828: 458:bow tie, bow-tie, bowtie
time: 9.754ms

此处我们需要先调用模型 1000 次，当然实际的时候不需要这么多次，这里只是为了把加载模型的时间排除在外

更改了一部分代码

import tensorflow as tf 变成 import tflite_runtime.interpreter as tflite
interpreter = tf.lite.Interpreter(model_path=args.model_file) 变成 interpreter = tflite.Interpreter(model_path=args.model_file)

0.919721: 653:military uniform
0.017762: 907:Windsor tie
0.007507: 668:mortarboard
0.005419: 466:bulletproof vest
0.003828: 458:bow tie, bow-tie, bowtie
time: 14.443ms

对比上面的 9.754ms 来说，好像变慢了，原因是什么？
而且更加的耗费 CPU 资源了，原因是？