How to change DataLoader in PyTorch to read one image for prediction?

Question

Currently, I have a pre-trained model that uses a DataLoader for reading a batch of images for training the model.

self.data_loader = torch.utils.data.DataLoader(dataset, batch_size=batch_size, shuffle=False, 
   num_workers=1, pin_memory=True)

...

model.eval()
for step, inputs in enumerate(test_loader.data_loader):
   outputs = model(torch.cat([inputs], 1))

...

I want to process (make predictions) on images, as they arrive from a queue. It should be similar to a code that reads a single image and runs the model to make predictions on it. Something along the following lines:

from PIL import Image

new_input = Image.open(image_path)
model.eval()
outputs = model(torch.cat([new_input ], 1))

I was wondering if you could guide me how to do this and apply the same transformations in the DataLoader.

This is really going to depend on how your `dataset` works. The details of which aren't provided in the question. — jodag, Apr 02 '20 at 22:06
@jodag I edited the question to provide more context. I appreciate your help. — Hamid R. Darabi, Apr 02 '20 at 23:52

score 1 · Accepted Answer · answered Apr 04 '20 at 18:33

You can use do it with IterableDataset :

from torch.utils.data import IterableDataset

class MyDataset(IterableDataset):
    def __init__(self, image_queue):
      self.queue = image_queue

    def read_next_image(self):
        while self.queue.qsize() > 0:
            # you can add transform here
            yield self.queue.get()
        return None

    def __iter__(self):
        return self.read_next_image()

and batch_size = 1 :

import queue
import torchvision.transforms.functional as TF

buffer = queue.Queue()
new_input = Image.open(image_path)
buffer.put(TF.to_tensor(new_input)) 
# ... Populate queue here

dataset = MyDataset(buffer)
dataloader = torch.utils.data.DataLoader(dataset, batch_size=1)
for data in dataloader:
   model(data) # data is one-image batch of size [1,3,H,W] where 3 - number of color channels

Thanks @Anton Ganichev – Hamid R. Darabi Apr 04 '20 at 20:10 — Hamid R. Darabi, Apr 04 '20 at 20:10

Marzi Heidari · Answer 2 · 2020-04-03T17:39:59.703

I don't know about dataLoader but you can load a single image using following function:

def safe_pil_loader(path, from_memory=False):
try:
    if from_memory:
        img = Image.open(path)
        res = img.convert('RGB')
    else:
        with open(path, 'rb') as f:
            img = Image.open(f)
            res = img.convert('RGB')
except:
    res = Image.new('RGB', (227, 227), color=0)
return res

And for applying transformation you can do as follows:

trans = transforms.Compose([
            transforms.Resize(299),
            transforms.CenterCrop(299),
            transforms.ToTensor(),
            normalize,
        ])
img=trans(img)

Thanks @Marzieh Heidari! – Hamid R. Darabi Apr 03 '20 at 17:25 — Hamid R. Darabi, Apr 03 '20 at 17:25

How to change DataLoader in PyTorch to read one image for prediction?

2 Answers2