How to predict own picture with a trained model？

1

I want to use my own trained model to predict the new picture, but the following problem has arisen. I don’t know where it is wrong, please help me.

Error message：File “/opt/anaconda3/lib/python3.6/site-packages/torch/nn/modules/conv.py”, line 301, in forward self.padding, self.dilation, self.groups)
RuntimeError: Expected 4-dimensional input for 4-dimensional weight [64, 3, 7, 7], but got input of size [1, 224, 224] instead

My code is like this：

Import needed packages

import torch
import torch.nn as nn
import torchvision.models as models
import numpy as np
from torch.autograd import Variable
from torchvision import datasets, transforms
import torch.nn.functional as F
import torchvision.utils as vutils
from io import open
import os
from PIL import Image

Load Data

data_transforms = transforms.Compose([
transforms.Resize(256),
transforms.CenterCrop(224),
transforms.ToTensor(),
transforms.Normalize([0.5,0.5,0.5], [0.5,0.5,0.5])
])

model_ft = torch.load(“model.pkl”)

imagepath = ‘./20170522/images1.1(074047-074310)/CPI__20170522_074049_909/001.png’
image = Image.open(imagepath)
imgblob = data_transforms(image)
imgblob.unsqueeze_(dim=0)
imgblob = Variable(imgblob)
torch.no_grad()
predict = F.softmax(model_ft(imgblob))
print(predict)

2

It looks like the channel dimension is missing in your input.
Could it be you’re loading a grayscale image?
If so, try to load your image as an RGB image using:

image = Image.open(imagepath).convert('RGB')
4

Your suggestion is very useful, thank you very much