Я новичок и учусь кодировать классификатор изображений. Моя цель - создать функцию predict
.
Любое предложение, чтобы исправить это?
В этом проекте я хочу использовать функцию прогнозирования для распознавания разных видов цветов. Чтобы я мог проверить их этикетки позже.
Попытка исправить: я уже использовал метод unsqueeze_(0)
и перешел от метода numpy к факелу. Я обычно получаю сообщение об ошибке, показанное ниже:
TypeError: img должно быть PIL
Код:
# Imports here
import pandas as pd
import numpy as np
import torch
from torch import nn
from torchvision import datasets, transforms, models
import torchvision.models as models
import torch.nn.functional as F
import torchvision.transforms.functional as F
from torch import optim
import json
from collections import OrderedDict
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
%matplotlib inline
from PIL import Image
def process_image(image):
#Scales, crops, and normalizes a PIL image for a PyTorch model,
#returns an Numpy array
# Process a PIL image for use in a PyTorch model
process = transforms.Compose([
transforms.Resize(256),
transforms.CenterCrop(224),
transforms.ToTensor(),
transforms.Normalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225])
])
image = process(image)
return image
# Predict
#Predict the class (or classes) of an image using a trained deep learning model.
def predict(image, model, topk=5):
img = process_image(image)
img = img.unsqueeze(0)
output = model.forward(img)
probs, labels = torch.topk(output, topk)
probs = probs.exp()
# Reverse the dict
idx_to_class = {val: key for key, val in model.class_to_idx.items()}
# Get the correct indices
top_classes = [idx_to_class[each] for each in classes]
return labels, probs
#Passing
probs, classes = predict(image, model)
print(probs)
print(classes)
Error
TypeError Traceback (most recent call last)
<ipython-input-92-b49fdcab5791> in <module>()
----> 1 probs, classes = predict(image, model)
2 print(probs)
3 print(classes)
<ipython-input-91-05809355bfe0> in predict(image, model, topk)
2 ‘’' Predict the class (or classes) of an image using a trained deep learning model.
3 ‘’'
----> 4 img = process_image(image)
5 img = img.unsqueeze(0)
6
<ipython-input-20-02663a696e34> in process_image(image)
11 std=[0.229, 0.224, 0.225])
12 ])
---> 13 image = process(image)
14 return image
/opt/conda/lib/python3.6/site-packages/torchvision-0.2.1-py3.6.egg/torchvision/transforms/transforms.py in __call__(self, img)
47 def __call__(self, img):
48 for t in self.transforms:
---> 49 img = t(img)
50 return img
51
/opt/conda/lib/python3.6/site-packages/torchvision-0.2.1-py3.6.egg/torchvision/transforms/transforms.py in __call__(self, img)
173 PIL Image: Rescaled image.
174 “”"
--> 175 return F.resize(img, self.size, self.interpolation)
176
177 def __repr__(self):
/opt/conda/lib/python3.6/site-packages/torchvision-0.2.1-py3.6.egg/torchvision/transforms/functional.py in resize(img, size, interpolation)
187 “”"
188 if not _is_pil_image(img):
--> 189 raise TypeError(‘img should be PIL Image. Got {}’.format(type(img)))
190 if not (isinstance(size, int) or (isinstance(size, collections.Iterable) and len(size) == 2)):
191 raise TypeError(‘Got inappropriate size arg: {}’.format(size))
TypeError: img should be PIL Image. Got <class ‘str’>
Все, чего я хочу, это получить такой же результат. Спасибо!
predict(image,model)
print(probs)
print(classes)
tensor([[ 0.5607, 0.3446, 0.0552, 0.0227, 0.0054]], device='cuda:0')
tensor([[ 8, 1, 31, 24, 7]], device='cuda:0')