안드로이드용 tflite image classification 모델 inference 스크립트

안녕하세요

저번 포스트에서 만들었던, 모델을 실제로 어떤 식으로 테스트 해볼수 있을까요?

 

다음 스크립트를 사용하면 테스트를 해볼 수 있습니다.

reference: https://www.tensorflow.org/lite/inference_with_metadata/task_library/image_classifier

 

#Imports

from tflite_support.task import vision
from tflite_support.task import core
from tflite_support.task import processor


from tensorflow_lite_support.python.task.core import base_options as base_options_module
from tensorflow_lite_support.python.task.processor.proto import classification_options_pb2

model_path = "models/this_model.tflite"
classification_options = classification_options_pb2.ClassificationOptions()
image_path = "this_img.jpg"


# Initialization
base_options = core.BaseOptions(file_name=model_path)
classification_options = processor.ClassificationOptions(max_results=2)
options = vision.ImageClassifierOptions(base_options=base_options, classification_options=classification_options)
classifier = vision.ImageClassifier.create_from_options(options)


# Run inference
image = vision.TensorImage.create_from_file(image_path)
classification_result = classifier.classify(image)


print("classification result : ", classification_result)