Feat2GS/metrics.py at main · fanegg/Feat2GS · GitHub

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
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
#
# Copyright (C) 2023, Inria
# GRAPHDECO research group, https://team.inria.fr/graphdeco
# All rights reserved.
#
# This software is free for non-commercial, research and evaluation use
# under the terms of the LICENSE.md file.
#
# For inquiries contact  george.drettakis@inria.fr
#

from pathlib import Path
import os
from PIL import Image
import torch
import torchvision.transforms.functional as tf
from utils.loss_utils import ssim, masked_ssim
from lpipsPyTorch import lpips
import json
from tqdm import tqdm
from utils.image_utils import psnr, masked_psnr
from argparse import ArgumentParser
import numpy as np
import matplotlib.pyplot as plt

def readImages(renders_dir, gt_dir, msk_dir):
    renders = []
    gts = []
    masks = []
    image_names = []
    for fname in os.listdir(renders_dir):
        render = Image.open(renders_dir / fname)
        gt = Image.open(gt_dir / fname)
        mask = Image.open(msk_dir / fname)
        renders.append(tf.to_tensor(render).unsqueeze(0)[:, :3, :, :].cuda())
        gts.append(tf.to_tensor(gt).unsqueeze(0)[:, :3, :, :].cuda())
        masks.append(tf.to_tensor(mask).int().unsqueeze(0)[:, 0, :, :].cuda())
        image_names.append(fname)
    return renders, gts, masks, image_names

def evaluate(args):

    full_dict = {}
    per_view_dict = {}

    print("")

    for scene_dir in args.model_paths:
        print("Scene:", scene_dir)
        full_dict[scene_dir] = {}
        per_view_dict[scene_dir] = {}

        test_dir = Path(scene_dir) / "test"

        for method in os.listdir(test_dir):
            print("Method:", method)

            full_dict[scene_dir][method] = {}
            per_view_dict[scene_dir][method] = {}

            method_dir = test_dir / method
            out_f = open(method_dir / 'metrics.txt', 'w')
            gt_dir = method_dir/ "gt"
            renders_dir = method_dir / "renders"
            msk_dir = method_dir/ "masks"
            renders, gts, masks, image_names = readImages(renders_dir, gt_dir, msk_dir)

            ssims = []
            psnrs = []
            lpipss = []

            for idx in tqdm(range(len(renders)), desc="Metric evaluation progress"):

                render_indx, gt_indx, mask_indx = renders[idx], gts[idx], masks[idx][:, None, ...]
                render_indx = render_indx * mask_indx
                gt_indx = gt_indx * mask_indx

                # s=masked_ssim(render_indx, gt_indx, mask_indx)
                # p=masked_psnr(render_indx, gt_indx, mask_indx)

                psnr_map = 20 * torch.log10(1.0 /( 1e-2 + torch.sqrt(((render_indx - gt_indx) ** 2).mean(1))))
                psnr_normalized = (psnr_map / 40).clamp(0, 1)
                psnr_image = (psnr_normalized.squeeze().cpu().numpy() * 255).astype(np.uint8)
                psnr_pil = Image.fromarray(psnr_image)
                psnr_pil.save(method_dir / f"psnr_map_{image_names[idx]}")

                ssim_map = ssim(render_indx, gt_indx, get_ssim_map=True).mean(1)
                ssim_normalized = ((ssim_map + 1)/ 2).clamp(0, 1)
                ssim_image = (ssim_normalized.squeeze().cpu().numpy() * 255).astype(np.uint8)
                ssim_pil = Image.fromarray(ssim_image)
                ssim_pil.save(method_dir / f"ssim_map_{image_names[idx]}")

                # mean square error
                jet_cmap = plt.cm.jet
                mse_map = ((render_indx - gt_indx) ** 2).mean(1)
                mse_map = mse_map.clamp(0, 0.2)*5
                mse_colored = jet_cmap(mse_map.squeeze().cpu().numpy())
                mse_image = (mse_colored[:, :, :3] * 255).astype(np.uint8)
                mse_pil = Image.fromarray(mse_image)
                mse_pil.save(method_dir / f"mse_map_{image_names[idx]}")

                # lpips
                lpips_map = lpips(render_indx, gt_indx, net_type='vgg', return_spatial_map=True)
                lpips_map = lpips_map.clamp(0, 1.0)
                lpips_colored = jet_cmap(lpips_map.squeeze().cpu().numpy())
                lpips_image = (lpips_colored[:, :, :3] * 255).astype(np.uint8)
                lpips_pil = Image.fromarray(lpips_image)
                lpips_pil.save(method_dir / f"lpips_map_{image_names[idx]}")

                s=ssim(render_indx, gt_indx)
                p=psnr(render_indx, gt_indx)
                l=lpips(render_indx, gt_indx, net_type='vgg')

                out_f.write(f"image name{image_names[idx]}, image idx: {idx}, PSNR: {p.item():.2f}, SSIM: {s:.4f}, LPIPS: {l.item():.4f}\n")
                ssims.append(s)
                psnrs.append(p)
                lpipss.append(l)

            print("  SSIM : {:>12.7f}".format(torch.tensor(ssims).mean(), ".5"))
            print("  PSNR : {:>12.7f}".format(torch.tensor(psnrs).mean(), ".5"))
            print("  LPIPS: {:>12.7f}".format(torch.tensor(lpipss).mean(), ".5"))
            print("")

            full_dict[scene_dir][method].update({"SSIM": torch.tensor(ssims).mean().item(),
                                                    "PSNR": torch.tensor(psnrs).mean().item(),
                                                    "LPIPS": torch.tensor(lpipss).mean().item()})
            per_view_dict[scene_dir][method].update({"SSIM": {name: ssim for ssim, name in zip(torch.tensor(ssims).tolist(), image_names)},
                                                        "PSNR": {name: psnr for psnr, name in zip(torch.tensor(psnrs).tolist(), image_names)},
                                                        "LPIPS": {name: lp for lp, name in zip(torch.tensor(lpipss).tolist(), image_names)}})

        with open(scene_dir + "/results.json", 'w') as fp:
            json.dump(full_dict[scene_dir], fp, indent=True)
        with open(scene_dir + "/per_view.json", 'w') as fp:
            json.dump(per_view_dict[scene_dir], fp, indent=True)


if __name__ == "__main__":
    device = torch.device("cuda:0")
    torch.cuda.set_device(device)

    # Set up command line argument parser
    parser = ArgumentParser(description="Training script parameters")
    parser.add_argument('--model_paths', '-m', required=True, nargs="+", type=str, default=[])
    parser.add_argument('--iteration', type=int, default=1000)
    parser.add_argument("--n_views", default=None, type=int)
    args = parser.parse_args()
    evaluate(args)