mnasnet.py

import math

import torch
import torch.nn as nn
from .utils import load_state_dict_from_url

__all__ = ['MNASNet', 'mnasnet0_5', 'mnasnet0_75', 'mnasnet1_0', 'mnasnet1_3']

_MODEL_URLS = {
    "mnasnet0_5":
    "https://download.pytorch.org/models/mnasnet0.5_top1_67.592-7c6cb539b9.pth",
    "mnasnet0_75": None,
    "mnasnet1_0":
    "https://download.pytorch.org/models/mnasnet1.0_top1_73.512-f206786ef8.pth",
    "mnasnet1_3": None
}

# Paper suggests 0.9997 momentum, for TensorFlow. Equivalent PyTorch momentum is
# 1.0 - tensorflow.
_BN_MOMENTUM = 1 - 0.9997


class _InvertedResidual(nn.Module):

    def __init__(self, in_ch, out_ch, kernel_size, stride, expansion_factor,
                 bn_momentum=0.1):
        super(_InvertedResidual, self).__init__()
        assert stride in [1, 2]
        assert kernel_size in [3, 5]
        mid_ch = in_ch * expansion_factor
        self.apply_residual = (in_ch == out_ch and stride == 1)
        self.layers = nn.Sequential(
            # Pointwise
            nn.Conv2d(in_ch, mid_ch, 1, bias=False),
            nn.BatchNorm2d(mid_ch, momentum=bn_momentum),
            nn.ReLU(inplace=True),
            # Depthwise
            nn.Conv2d(mid_ch, mid_ch, kernel_size, padding=kernel_size // 2,
                      stride=stride, groups=mid_ch, bias=False),
            nn.BatchNorm2d(mid_ch, momentum=bn_momentum),
            nn.ReLU(inplace=True),
            # Linear pointwise. Note that there's no activation.
            nn.Conv2d(mid_ch, out_ch, 1, bias=False),
            nn.BatchNorm2d(out_ch, momentum=bn_momentum))

    def forward(self, input):
        if self.apply_residual:
            return self.layers(input) + input
        else:
            return self.layers(input)


def _stack(in_ch, out_ch, kernel_size, stride, exp_factor, repeats,
           bn_momentum):
    """ Creates a stack of inverted residuals. """
    assert repeats >= 1
    # First one has no skip, because feature map size changes.
    first = _InvertedResidual(in_ch, out_ch, kernel_size, stride, exp_factor,
                              bn_momentum=bn_momentum)
    remaining = []
    for _ in range(1, repeats):
        remaining.append(
            _InvertedResidual(out_ch, out_ch, kernel_size, 1, exp_factor,
                              bn_momentum=bn_momentum))
    return nn.Sequential(first, *remaining)


def _round_to_multiple_of(val, divisor, round_up_bias=0.9):
    """ Asymmetric rounding to make `val` divisible by `divisor`. With default
    bias, will round up, unless the number is no more than 10% greater than the
    smaller divisible value, i.e. (83, 8) -> 80, but (84, 8) -> 88. """
    assert 0.0 < round_up_bias < 1.0
    new_val = max(divisor, int(val + divisor / 2) // divisor * divisor)
    return new_val if new_val >= round_up_bias * val else new_val + divisor


def _scale_depths(depths, alpha):
    """ Scales tensor depths as in reference MobileNet code, prefers rouding up
    rather than down. """
    return [_round_to_multiple_of(depth * alpha, 8) for depth in depths]


class MNASNet(torch.nn.Module):
    """ MNASNet, as described in https://arxiv.org/pdf/1807.11626.pdf.
    >>> model = MNASNet(1000, 1.0)
    >>> x = torch.rand(1, 3, 224, 224)
    >>> y = model(x)
    >>> y.dim()
    1
    >>> y.nelement()
    1000
    """

    def __init__(self, alpha, num_classes=1000, dropout=0.2):
        super(MNASNet, self).__init__()
        depths = _scale_depths([24, 40, 80, 96, 192, 320], alpha)
        layers = [
            # First layer: regular conv.
            nn.Conv2d(3, 32, 3, padding=1, stride=2, bias=False),
            nn.BatchNorm2d(32, momentum=_BN_MOMENTUM),
            nn.ReLU(inplace=True),
            # Depthwise separable, no skip.
            nn.Conv2d(32, 32, 3, padding=1, stride=1, groups=32, bias=False),
            nn.BatchNorm2d(32, momentum=_BN_MOMENTUM),
            nn.ReLU(inplace=True),
            nn.Conv2d(32, 16, 1, padding=0, stride=1, bias=False),
            nn.BatchNorm2d(16, momentum=_BN_MOMENTUM),
            # MNASNet blocks: stacks of inverted residuals.
            _stack(16, depths[0], 3, 2, 3, 3, _BN_MOMENTUM),
            _stack(depths[0], depths[1], 5, 2, 3, 3, _BN_MOMENTUM),
            _stack(depths[1], depths[2], 5, 2, 6, 3, _BN_MOMENTUM),
            _stack(depths[2], depths[3], 3, 1, 6, 2, _BN_MOMENTUM),
            _stack(depths[3], depths[4], 5, 2, 6, 4, _BN_MOMENTUM),
            _stack(depths[4], depths[5], 3, 1, 6, 1, _BN_MOMENTUM),
            # Final mapping to classifier input.
            nn.Conv2d(depths[5], 1280, 1, padding=0, stride=1, bias=False),
            nn.BatchNorm2d(1280, momentum=_BN_MOMENTUM),
            nn.ReLU(inplace=True),
        ]
        self.layers = nn.Sequential(*layers)
        self.classifier = nn.Sequential(nn.Dropout(p=dropout, inplace=True),
                                        nn.Linear(1280, num_classes))
        self._initialize_weights()

    def forward(self, x):
        x = self.layers(x)
        # Equivalent to global avgpool and removing H and W dimensions.
        x = x.mean([2, 3])
        return self.classifier(x)

    def _initialize_weights(self):
        for m in self.modules():
            if isinstance(m, nn.Conv2d):
                nn.init.kaiming_normal_(m.weight, mode="fan_out",
                                        nonlinearity="relu")
                if m.bias is not None:
                    nn.init.zeros_(m.bias)
            elif isinstance(m, nn.BatchNorm2d):
                nn.init.ones_(m.weight)
                nn.init.zeros_(m.bias)
            elif isinstance(m, nn.Linear):
                nn.init.normal_(m.weight, 0.01)
                nn.init.zeros_(m.bias)


def _load_pretrained(model_name, model, progress):
    if model_name not in _MODEL_URLS or _MODEL_URLS[model_name] is None:
        raise ValueError(
            "No checkpoint is available for model type {}".format(model_name))
    checkpoint_url = _MODEL_URLS[model_name]
    model.load_state_dict(load_state_dict_from_url(checkpoint_url, progress=progress))


def mnasnet0_5(pretrained=False, progress=True, **kwargs):
    """MNASNet with depth multiplier of 0.5 from
    `"MnasNet: Platform-Aware Neural Architecture Search for Mobile"
    <https://arxiv.org/pdf/1807.11626.pdf>`_.
    Args:
        pretrained (bool): If True, returns a model pre-trained on ImageNet
        progress (bool): If True, displays a progress bar of the download to stderr
    """
    model = MNASNet(0.5, **kwargs)
    if pretrained:
        _load_pretrained("mnasnet0_5", model, progress)
    return model


def mnasnet0_75(pretrained=False, progress=True, **kwargs):
    """MNASNet with depth multiplier of 0.75 from
    `"MnasNet: Platform-Aware Neural Architecture Search for Mobile"
    <https://arxiv.org/pdf/1807.11626.pdf>`_.
    Args:
        pretrained (bool): If True, returns a model pre-trained on ImageNet
        progress (bool): If True, displays a progress bar of the download to stderr
    """
    model = MNASNet(0.75, **kwargs)
    if pretrained:
        _load_pretrained("mnasnet0_75", model, progress)
    return model


def mnasnet1_0(pretrained=False, progress=True, **kwargs):
    """MNASNet with depth multiplier of 1.0 from
    `"MnasNet: Platform-Aware Neural Architecture Search for Mobile"
    <https://arxiv.org/pdf/1807.11626.pdf>`_.
    Args:
        pretrained (bool): If True, returns a model pre-trained on ImageNet
        progress (bool): If True, displays a progress bar of the download to stderr
    """
    model = MNASNet(1.0, **kwargs)
    if pretrained:
        _load_pretrained("mnasnet1_0", model, progress)
    return model


def mnasnet1_3(pretrained=False, progress=True, **kwargs):
    """MNASNet with depth multiplier of 1.3 from
    `"MnasNet: Platform-Aware Neural Architecture Search for Mobile"
    <https://arxiv.org/pdf/1807.11626.pdf>`_.
    Args:
        pretrained (bool): If True, returns a model pre-trained on ImageNet
        progress (bool): If True, displays a progress bar of the download to stderr
    """
    model = MNASNet(1.3, **kwargs)
    if pretrained:
        _load_pretrained("mnasnet1_3", model, progress)
    return model