You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

81 lines
2.6 KiB

!pip install paddlehub==1.8.1 -i https://pypi.tuna.tsinghua.edu.cn/simple
import paddlehub as hub
module = hub.Module(name="mobilenet_v2_imagenet")
!unzip -o /data/shixunfiles/26a2e3c3b2c50fe54e2fcab6e031a141_1607408726958.zip
from paddlehub.dataset.base_cv_dataset import BaseCVDataset
class DemoDataset(BaseCVDataset):
def __init__(self):
self.dataset_dir = "car_datasets"
super(DemoDataset, self).__init__(
base_path=self.dataset_dir,
train_list_file="train_list.txt",
validate_list_file="validate_list.txt",
test_list_file="test_list.txt",
label_list_file="label_list.txt",
)
dataset = DemoDataset()
data_reader = hub.reader.ImageClassificationReader(
image_width=module.get_expected_image_width(),
image_height=module.get_expected_image_height(),
images_mean=module.get_pretrained_images_mean(),
images_std=module.get_pretrained_images_std(),
dataset=dataset)
config = hub.RunConfig(
use_cuda=False,
num_epoch=10,
batch_size=32,
eval_interval=50,
strategy=hub.finetune.strategy.DefaultFinetuneStrategy())
input_dict, output_dict, program = module.context(trainable=True)
img = input_dict["image"]
feature_map = output_dict["feature_map"]
feed_list = [img.name]
task = hub.ImageClassifierTask(
data_reader=data_reader,
feed_list=feed_list,
feature=feature_map,
num_classes=dataset.num_labels,
config=config)
run_states = task.finetune_and_eval()
import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
from pandas import Series,DataFrame
%matplotlib inline
import os
dirs=os.listdir('car_datasets/test')
num=0
for i in os.listdir('car_datasets/test'):
m='car_datasets/test/'+i
dirs[num]=m
num+=1
s=0
b=0
a=os.listdir('car_datasets/test')
for i in a:
b+=len(os.listdir('car_datasets/test/'+i))
data=[]
for i in range(b):
data.append('o')
for i in dirs:
for j in os.listdir(i):
n=i+'/'+j
data[s]=n
s+=1
label_map = dataset.label_dict()
index = 0
true=0
run_states = task.predict(data=data)
results = [run_state.run_results for run_state in run_states]
for batch_result in results:
batch_result = np.argmax(batch_result, axis=2)[0]
for result in batch_result:
index += 1
result = label_map[result]
actual=os.path.dirname(data[index - 1])
actual=actual.split('/')
if actual[-1]==result:
true+=1
print("input %i is %s, and the predict result is ( %s )" %
(index, data[index - 1], result))
print( '预测正确率为{:.2%}'.format(true/index))