metrics.py

import numpy as np
import torch

from scipy import linalg
from scipy.stats import entropy
from torchvision.models.inception import inception_v3
from torch import nn
from torch.autograd import Variable
from torch.nn import functional as F

EPSILON = 1e-20

class Score:
    @staticmethod
    def inception_score(imgs, cuda=True, batch_size=32, resize=True, splits=1):
        """Computes the inception score of the generated images imgs
        imgs -- Torch dataset of (3xHxW) numpy images normalized in the range [-1, 1]
        cuda -- whether or not to run on GPU
        batch_size -- batch size for feeding into Inception v3
        splits -- number of splits
        """
        N = len(imgs)

        assert batch_size > 0
        # assert N > batch_size

        # Set up dtype
        if cuda:
            dtype = torch.cuda.FloatTensor
        else:
            if torch.cuda.is_available():
                print("WARNING: You have a CUDA device, so you should probably set cuda=True")
            dtype = torch.FloatTensor

        # Set up dataloader
        dataloader = torch.utils.data.DataLoader(imgs, batch_size=batch_size)

        # Load inception model
        inception_model = inception_v3(pretrained=True, transform_input=False).type(dtype)
        inception_model.eval()
        up = nn.Upsample(size=(299, 299), mode='bilinear').type(dtype)

        def get_pred(x):
            if resize:
                x = up(x)
            x = inception_model(x)
            return F.softmax(x).data.cpu().numpy()

        # Get predictions
        preds = np.zeros((N, 1000))

        for i, batch in enumerate(dataloader, 0):
            batch = batch.type(dtype)
            batchv = Variable(batch)
            batch_size_i = batch.size()[0]

            preds[i * batch_size:i * batch_size + batch_size_i] = get_pred(batchv)

        # Now compute the mean kl-div
        split_scores = []

        for k in range(splits):
            part = preds[k * (N // splits): (k + 1) * (N // splits), :]
            py = np.mean(part, axis=0)
            scores = []
            for i in range(part.shape[0]):
                pyx = part[i, :]
                scores.append(entropy(pyx, py))
            split_scores.append(np.exp(np.mean(scores)))

        return np.mean(split_scores), np.std(split_scores)

    @staticmethod
    def frechet_inception_distance(real_data, fake_data):
        real_data = real_data.permute(0, 2, 3, 1)
        fake_data = fake_data.permute(0, 2, 3, 1)
        if isinstance(real_data, torch.autograd.Variable):
            real_np = real_data.data.cpu().numpy()
        if isinstance(fake_data, torch.autograd.Variable):
            fake_np = fake_data.data.cpu().numpy()

        scores = []
        for id, data in enumerate(real_np):
            m = real_np[id].mean(0)
            m_w = fake_np[id].mean(0)

            C = np.cov(real_np[id].transpose())
            C_w = np.cov(fake_np[id].transpose())
            C_C_w_sqrt = linalg.sqrtm(C.dot(C_w), True).real

            score = m.dot(m) + m_w.dot(m_w) - 2 * m_w.dot(m) + \
                    np.trace(C + C_w - 2 * C_C_w_sqrt)

            scores.append(np.sqrt(score))

        return np.mean(scores)