File size: 2,722 Bytes
7df68a5
 
0aff08c
7df68a5
 
 
 
 
 
 
 
0aff08c
09a89dc
0aff08c
7df68a5
 
 
46438b0
09a89dc
 
7df68a5
 
09a89dc
 
 
7df68a5
 
 
 
 
 
 
 
 
09a89dc
7df68a5
 
09a89dc
7df68a5
 
 
 
 
09a89dc
 
7df68a5
 
 
 
09a89dc
 
7df68a5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
09a89dc
7df68a5
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
from __future__ import annotations

import pathlib
import pickle
import sys

import numpy as np
import torch
import torch.nn as nn
from huggingface_hub import hf_hub_download

app_dir = pathlib.Path(__file__).parent
submodule_dir = app_dir / "StyleGAN-Human"
sys.path.insert(0, submodule_dir.as_posix())


class Model:
    def __init__(self):
        self.device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
        self.model = self.load_model("stylegan_human_v2_1024.pkl")

    def load_model(self, file_name: str) -> nn.Module:
        path = hf_hub_download("public-data/StyleGAN-Human", f"models/{file_name}")
        with open(path, "rb") as f:
            model = pickle.load(f)["G_ema"]
        model.eval()
        model.to(self.device)
        with torch.inference_mode():
            z = torch.zeros((1, model.z_dim)).to(self.device)
            label = torch.zeros([1, model.c_dim], device=self.device)
            model(z, label, force_fp32=True)
        return model

    def generate_z(self, z_dim: int, seed: int) -> torch.Tensor:
        return torch.from_numpy(np.random.RandomState(seed).randn(1, z_dim)).to(self.device).float()

    @torch.inference_mode()
    def generate_single_image(self, seed: int, truncation_psi: float) -> np.ndarray:
        seed = int(np.clip(seed, 0, np.iinfo(np.uint32).max))

        z = self.generate_z(self.model.z_dim, seed)
        label = torch.zeros([1, self.model.c_dim], device=self.device)

        out = self.model(z, label, truncation_psi=truncation_psi, force_fp32=True)
        out = (out.permute(0, 2, 3, 1) * 127.5 + 128).clamp(0, 255).to(torch.uint8)
        return out[0].cpu().numpy()

    @torch.inference_mode()
    def generate_interpolated_images(
        self, seed0: int, psi0: float, seed1: int, psi1: float, num_intermediate: int
    ) -> list[np.ndarray]:
        seed0 = int(np.clip(seed0, 0, np.iinfo(np.uint32).max))
        seed1 = int(np.clip(seed1, 0, np.iinfo(np.uint32).max))

        z0 = self.generate_z(self.model.z_dim, seed0)
        z1 = self.generate_z(self.model.z_dim, seed1)
        vec = z1 - z0
        dvec = vec / (num_intermediate + 1)
        zs = [z0 + dvec * i for i in range(num_intermediate + 2)]
        dpsi = (psi1 - psi0) / (num_intermediate + 1)
        psis = [psi0 + dpsi * i for i in range(num_intermediate + 2)]

        label = torch.zeros([1, self.model.c_dim], device=self.device)

        res = []
        for z, psi in zip(zs, psis):
            out = self.model(z, label, truncation_psi=psi, force_fp32=True)
            out = (out.permute(0, 2, 3, 1) * 127.5 + 128).clamp(0, 255).to(torch.uint8)
            out = out[0].cpu().numpy()
            res.append(out)
        return res