evaluate.py

from models import models_vit
import sys
import argparse
import os
import time
import shutil
import torch
import torch.nn as nn
import torch.nn.parallel
import torch.backends.cudnn as cudnn
import torch.optim
import torch.utils.data
import torch.utils.data.distributed
from models.Generate_Model import GenerateModel
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt
import numpy as np
import itertools
import datetime
from dataloader.video_dataloader import train_data_loader, test_data_loader
from sklearn.metrics import confusion_matrix
import tqdm
import warnings
warnings.filterwarnings("ignore", category=UserWarning)
import random

seed = 1
random.seed(seed)  
np.random.seed(seed) 
torch.manual_seed(seed)
torch.cuda.manual_seed(seed)
torch.cuda.manual_seed_all(seed)

def parse_args():
    parser = argparse.ArgumentParser()
    parser.add_argument('--dataset', type=str, default='DFEW')
    parser.add_argument('--checkpoint', type=str)
    parser.add_argument('--temporal-layers', type=int, default=1)
    parser.add_argument('--img-size', type=int, default=224)

    parser.add_argument('--workers', type=int, default=8)

    parser.add_argument('--fold', type=int, default=1)
    args = parser.parse_args()
    return args

def main(set, args):
    
    data_set = set
    
    if args.dataset == "DFEW":
        print("*********** DFEW Dataset Fold  " + str(data_set) + " ***********")
        test_annotation_file_path = "./annotation/DFEW_set_"+str(data_set)+"_test.txt"
        args.number_class = 7     
    elif args.dataset == "MAFW":
        print("*********** MAFW Dataset Fold  " + str(data_set) + " ***********")
        test_annotation_file_path = "./annotation/MAFW_set_"+str(data_set)+"_test_faces.txt"
        args.number_class = 11      
    model = GenerateModel(args=args)
    model = torch.nn.DataParallel(model).cuda()
    test_data = test_data_loader(list_file=test_annotation_file_path,
                                 num_segments=16,
                                 duration=1,
                                 image_size=args.img_size)

    val_loader = torch.utils.data.DataLoader(test_data,
                                             batch_size=1, #args.batch_size,
                                             shuffle=False,
                                             num_workers=args.workers,
                                             pin_memory=True)
 
    uar, war = computer_uar_war(val_loader, model, args.checkpoint, data_set)
  
    return uar, war

class AverageMeter(object):
    """Computes and stores the average and current value"""
    def __init__(self, name, fmt=':f'):
        self.name = name
        self.fmt = fmt
        self.reset()

    def reset(self):
        self.val = 0
        self.avg = 0
        self.sum = 0
        self.count = 0

    def update(self, val, n=1):
        self.val = val
        self.sum += val * n
        self.count += n
        self.avg = self.sum / self.count

    def __str__(self):
        fmtstr = '{name} {val' + self.fmt + '} ({avg' + self.fmt + '})'
        return fmtstr.format(**self.__dict__)


class ProgressMeter(object):
    def __init__(self, num_batches, meters, prefix="", log_txt_path=""):
        self.batch_fmtstr = self._get_batch_fmtstr(num_batches)
        self.meters = meters
        self.prefix = prefix
        self.log_txt_path = log_txt_path

    def display(self, batch):
        entries = [self.prefix + self.batch_fmtstr.format(batch)]
        entries += [str(meter) for meter in self.meters]
        print_txt = '\t'.join(entries)
        print(print_txt)
        with open(self.log_txt_path, 'a') as f:
            f.write(print_txt + '\n')

    def _get_batch_fmtstr(self, num_batches):
        num_digits = len(str(num_batches // 1))
        fmt = '{:' + str(num_digits) + 'd}'
        return '[' + fmt + '/' + fmt.format(num_batches) + ']'


def accuracy(output, target, topk=(1,)):
    """Computes the accuracy over the k top predictions for the specified values of k"""
    with torch.no_grad():
        maxk = max(topk)
        batch_size = target.size(0)
        _, pred = output.topk(maxk, 1, True, True)
        pred = pred.t()
        correct = pred.eq(target.view(1, -1).expand_as(pred))
        res = []
        for k in topk:
            correct_k = correct[:k].contiguous().view(-1).float().sum(0, keepdim=True)
            res.append(correct_k.mul_(100.0 / batch_size))
        return res


class RecorderMeter(object):
    """Computes and stores the minimum loss value and its epoch index"""
    def __init__(self, total_epoch):
        self.reset(total_epoch)

    def reset(self, total_epoch):
        self.total_epoch = total_epoch
        self.current_epoch = 0
        self.epoch_losses = np.zeros((self.total_epoch, 2), dtype=np.float32)    # [epoch, train/val]
        self.epoch_accuracy = np.zeros((self.total_epoch, 2), dtype=np.float32)  # [epoch, train/val]

    def update(self, idx, train_loss, train_acc, val_loss, val_acc):
        self.epoch_losses[idx, 0] = train_loss * 50
        self.epoch_losses[idx, 1] = val_loss * 50
        self.epoch_accuracy[idx, 0] = train_acc
        self.epoch_accuracy[idx, 1] = val_acc
        self.current_epoch = idx + 1

    def plot_curve(self, save_path):

        title = 'the accuracy/loss curve of train/val'
        dpi = 80
        width, height = 1600, 800
        legend_fontsize = 10
        figsize = width / float(dpi), height / float(dpi)

        fig = plt.figure(figsize=figsize)
        x_axis = np.array([i for i in range(self.total_epoch)])  # epochs
        y_axis = np.zeros(self.total_epoch)

        plt.xlim(0, self.total_epoch)
        plt.ylim(0, 100)
        interval_y = 5
        interval_x = 1
        plt.xticks(np.arange(0, self.total_epoch + interval_x, interval_x))
        plt.yticks(np.arange(0, 100 + interval_y, interval_y))
        plt.grid()
        plt.title(title, fontsize=20)
        plt.xlabel('the training epoch', fontsize=16)
        plt.ylabel('accuracy', fontsize=16)

        y_axis[:] = self.epoch_accuracy[:, 0]
        plt.plot(x_axis, y_axis, color='g', linestyle='-', label='train-accuracy', lw=2)
        plt.legend(loc=4, fontsize=legend_fontsize)

        y_axis[:] = self.epoch_accuracy[:, 1]
        plt.plot(x_axis, y_axis, color='y', linestyle='-', label='valid-accuracy', lw=2)
        plt.legend(loc=4, fontsize=legend_fontsize)

        y_axis[:] = self.epoch_losses[:, 0]
        plt.plot(x_axis, y_axis, color='g', linestyle=':', label='train-loss-x50', lw=2)
        plt.legend(loc=4, fontsize=legend_fontsize)

        y_axis[:] = self.epoch_losses[:, 1]
        plt.plot(x_axis, y_axis, color='y', linestyle=':', label='valid-loss-x50', lw=2)
        plt.legend(loc=4, fontsize=legend_fontsize)

        if save_path is not None:
            fig.savefig(save_path, dpi=dpi, bbox_inches='tight')
            # print('Curve was saved')
        plt.close(fig)


def plot_confusion_matrix(cm, classes, normalize=False, title='confusion matrix', cmap=plt.cm.Blues):
    """
    This function prints and plots the confusion matrix.
    Normalization can be applied by setting `normalize=True`.
    """
    plt.imshow(cm, interpolation='nearest', cmap=cmap)
    plt.title(title, fontsize=16)
    plt.colorbar()
    tick_marks = np.arange(len(classes))
    plt.xticks(tick_marks, classes, rotation=45)
    plt.yticks(tick_marks, classes)

    fmt = '.2f' if normalize else 'd'
    thresh = cm.max() / 2.
    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):
        plt.text(j, i, format(cm[i, j], fmt), fontsize=12,
                 horizontalalignment="center",
                 color="white" if cm[i, j] > thresh else "black")

    plt.ylabel('True label', fontsize=18)
    plt.xlabel('Predicted label', fontsize=18)
    plt.tight_layout()

def computer_uar_war(val_loader, model, checkpoint_path, data_set):
    
    pre_trained_dict = torch.load(checkpoint_path)['state_dict']
    model.load_state_dict(pre_trained_dict)
    model.eval()

    correct = 0
    with torch.no_grad():
        for i, (images, target, audio) in enumerate(tqdm.tqdm(val_loader)):
            
            images = images.cuda()
            target = target.cuda()
            audio = audio.cuda()
            output = model(images, audio)        

            predicted = output.argmax(dim=1, keepdim=True)
            correct += predicted.eq(target.view_as(predicted)).sum().item()

            if i == 0:
                all_predicted = predicted
                all_targets = target
            else:
                all_predicted = torch.cat((all_predicted, predicted), 0)
                all_targets = torch.cat((all_targets, target), 0)

    war = 100. * correct / len(val_loader.dataset)
    
    # Compute confusion matrix
    _confusion_matrix = confusion_matrix(all_targets.data.cpu().numpy(), all_predicted.cpu().numpy())
    np.set_printoptions(precision=4)
    normalized_cm = _confusion_matrix.astype('float') / _confusion_matrix.sum(axis=1)[:, np.newaxis]
    normalized_cm = normalized_cm * 100
    list_diag = np.diag(normalized_cm)
    uar = list_diag.mean()
        
    print("Confusion Matrix Diag:", list_diag)
    print("UAR: %0.2f" % uar)
    print("WAR: %0.2f" % war)
    return uar, war


if __name__ == '__main__':
    args = parse_args() 
    print('************************')
    for k, v in vars(args).items():
        print(k,'=',v)
    print('************************')
    uar, war = main(args.fold, args)
       
    print('********* Final Results *********')   
    print("UAR: %0.2f" % (uar))
    print("WAR: %0.2f" % (war))
    print('*********************************')