predict.py

import os
import torch
from PIL import Image
from torchvision import transforms
import matplotlib.pyplot as plt
from torchstat import stat
# import torchvision.models as models

# from model import mobile_vit_xx_small as create_model
# from model import mobile_vit_x_small as create_model
from model import mobile_vit_small as create_model


def main():
    device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
    # print(torch.cuda.is_available())

    img_size = 224
    data_transform = transforms.Compose(
        [transforms.Resize(int(img_size * 1.14)),
         transforms.CenterCrop(img_size),
         transforms.ToTensor(),
         transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])])

    # load image
    img_path = r"test_data/pic(1).JPEG"
    assert os.path.exists(img_path), "file: '{}' dose not exist.".format(img_path)
    img = Image.open(img_path)
    plt.imshow(img)
    # [N, C, H, W]
    img = data_transform(img)
    # expand batch dimension
    img = torch.unsqueeze(img, dim=0)
    #
    # # read class_indict
    # json_path = './class_indices.json'
    # assert os.path.exists(json_path), "file: '{}' dose not exist.".format(json_path)
    #
    # with open(json_path, "r") as f:
    #     class_indict = json.load(f)

    # create model
    model = create_model(num_classes=1000).to(device)
    # load model weights
    model_weight_path = r"./weights/mobilevit_s.pt"
    # model_weight_path = r"./weights/mobilevit_xs.pt"
    # model_weight_path = r"./weights/mobilevit_xxs.pt"
    model.load_state_dict(torch.load(model_weight_path, map_location=device))
    model.eval()

    print(img.to(device))

    print(stat(model.to("cpu"),(3, 224, 224)))

    with torch.no_grad():
        # predict class
        output = torch.squeeze(model(img.to(device))).cpu()
        predict = torch.softmax(output, dim=0)
        predict_cla = torch.argmax(predict).numpy()

    # print(predict)
    print(predict_cla)

    # print_res = "class: {}   prob: {:.3}".format(class_indict[str(predict_cla)],
    #                                              predict[predict_cla].numpy())
    # plt.title(print_res)
    # for i in range(len(predict)):
    #     print("class: {:10}   prob: {:.3}".format(class_indict[str(i)],
    #                                               predict[i].numpy()))
    # plt.show()


if __name__ == '__main__':
    main()