generate_w.py

# Copyright (c) 2021, NVIDIA CORPORATION.  All rights reserved.
#
# NVIDIA CORPORATION and its licensors retain all intellectual property
# and proprietary rights in and to this software, related documentation
# and any modifications thereto.  Any use, reproduction, disclosure or
# distribution of this software and related documentation without an express
# license agreement from NVIDIA CORPORATION is strictly prohibited.

"""Generate images using pretrained network pickle."""

import os
import re
from typing import List, Optional
import torchvision
from torchvision.transforms import Compose, Resize, CenterCrop, ToTensor, Normalize
import click
import dnnlib
import numpy as np
import PIL.Image
import torch
from torch import linalg as LA
import clip
from PIL import Image
import legacy
import torch.nn.functional as F
import cv2
import matplotlib.pyplot as plt
from torch_utils import misc
from torch_utils import persistence
from torch_utils.ops import conv2d_resample
from torch_utils.ops import upfirdn2d
from torch_utils.ops import bias_act
from torch_utils.ops import fma
import random
import math
import time
import id_loss


def block_forward(self, x, img, ws, shapes, force_fp32=False, fused_modconv=None, **layer_kwargs):
        misc.assert_shape(ws, [None, self.num_conv + self.num_torgb, self.w_dim])
        w_iter = iter(ws.unbind(dim=1))
        dtype = torch.float16 if self.use_fp16 and not force_fp32 else torch.float32
        memory_format = torch.channels_last if self.channels_last and not force_fp32 else torch.contiguous_format
        if fused_modconv is None:
            with misc.suppress_tracer_warnings(): # this value will be treated as a constant
                fused_modconv = (not self.training) and (dtype == torch.float32 or int(x.shape[0]) == 1)

        # Input.
        if self.in_channels == 0:
            x = self.const.to(dtype=dtype, memory_format=memory_format)
            x = x.unsqueeze(0).repeat([ws.shape[0], 1, 1, 1])
        else:
            misc.assert_shape(x, [None, self.in_channels, self.resolution // 2, self.resolution // 2])
            x = x.to(dtype=dtype, memory_format=memory_format)

        # Main layers.
        if self.in_channels == 0:
            x = self.conv1(x, next(w_iter)[...,:shapes[0]], fused_modconv=fused_modconv, **layer_kwargs)
        elif self.architecture == 'resnet':
            y = self.skip(x, gain=np.sqrt(0.5))
            x = self.conv0(x, next(w_iter), fused_modconv=fused_modconv, **layer_kwargs)
            x = self.conv1(x, next(w_iter), fused_modconv=fused_modconv, gain=np.sqrt(0.5), **layer_kwargs)
            x = y.add_(x)
        else:
            x = self.conv0(x, next(w_iter)[...,:shapes[0]], fused_modconv=fused_modconv, **layer_kwargs)
            x = self.conv1(x, next(w_iter)[...,:shapes[1]], fused_modconv=fused_modconv, **layer_kwargs)

        # ToRGB.
        if img is not None:
            misc.assert_shape(img, [None, self.img_channels, self.resolution // 2, self.resolution // 2])
            img = upfirdn2d.upsample2d(img, self.resample_filter)
        if self.is_last or self.architecture == 'skip':
            y = self.torgb(x, next(w_iter)[...,:shapes[2]], fused_modconv=fused_modconv)
            y = y.to(dtype=torch.float32, memory_format=torch.contiguous_format)
            img = img.add_(y) if img is not None else y

        assert x.dtype == dtype
        assert img is None or img.dtype == torch.float32
        return x, img

def unravel_index(index, shape):
    out = []
    for dim in reversed(shape):
        out.append(index % dim)
        index = index // dim
    return tuple(reversed(out))


def num_range(s: str) -> List[int]:
    '''Accept either a comma separated list of numbers 'a,b,c' or a range 'a-c' and return as a list of ints.'''

    range_re = re.compile(r'^(\d+)-(\d+)$')
    m = range_re.match(s)
    if m:
        return list(range(int(m.group(1)), int(m.group(2))+1))
    vals = s.split(',')
    return [int(x) for x in vals]


@click.command()
@click.pass_context
@click.option('--network', 'network_pkl', help='Network pickle filename', required=True)
@click.option('--seeds', type=num_range, help='List of random seeds')
@click.option('--trunc', 'truncation_psi', type=float, help='Truncation psi', default=1, show_default=True)
@click.option('--class', 'class_idx', type=int, help='Class label (unconditional if not specified)')
@click.option('--noise-mode', help='Noise mode', type=click.Choice(['const', 'random', 'none']), default='const', show_default=True)
def generate_images(
    ctx: click.Context,
    network_pkl: str,
    seeds: Optional[List[int]],
    truncation_psi: float,
    noise_mode: str,
    class_idx: Optional[int],
    projected_w: Optional[str],
    projected_s: Optional[str]
):

    print('Loading networks from "%s"...' % network_pkl)
    # Use GPU if available
    if torch.cuda.is_available():
        device = torch.device("cuda")
    else:
        device = torch.device("cpu")

    with dnnlib.util.open_url(network_pkl) as f:
        G = legacy.load_network_pkl(f)['G_ema'].to(device) # type: ignore

    if seeds is None:
        ctx.fail('--seeds option is required when not using --projected-w')

    # Labels.
    label = torch.zeros([1, G.c_dim], device=device).requires_grad_()
    if G.c_dim != 0:
        if class_idx is None:
            ctx.fail('Must specify class label with --class when using a conditional network')
        label[:, class_idx] = 1
    else:
        if class_idx is not None:
            print ('warn: --class=lbl ignored when running on an unconditional network')

    z = torch.from_numpy(np.random.RandomState(seeds[0]).randn(1, G.z_dim)).to(device)
    ws = G.mapping(z, label, truncation_psi=truncation_psi)
    np.savez(f'encoder4editing/projected_w.npz', w=ws.detach().cpu().numpy())


if __name__ == "__main__":
    generate_images()