-
Notifications
You must be signed in to change notification settings - Fork 9
/
Copy pathdemo-torch.py
299 lines (274 loc) · 9.37 KB
/
demo-torch.py
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
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
from timeit import timeit
import torch
from torchvision.transforms.functional import to_tensor
from PIL import Image
import pixeloe.torch.env as pixeloe_env
from pixeloe.torch.pixelize import pixelize
from pixeloe.torch.outline import outline_expansion
from pixeloe.torch.utils import to_numpy, pre_resize
from pixeloe.torch.minmax import dilate_cont, erode_cont, KERNELS
DEVICE = "cuda"
DTYPE = torch.float16
COMPILE = True
N = 200
if __name__ == "__main__":
img = Image.open("./img/snow-leopard.webp")
pixeloe_env.TORCH_COMPILE = COMPILE
img_t = to_tensor(img).to(DEVICE).to(DTYPE)[None]
oe_t, w = outline_expansion(img_t, 6, 6, 8, 10, 3)
oe = Image.fromarray(to_numpy(oe_t)[0])
oe.save("./img/snow-leopard-oe-orig.webp", lossless=True, quality=0)
pixeloe_env.TORCH_COMPILE = False
patch_size = 5
target_size = 240
thickness = 4
lg_patch_size = 4
lg_target_size = 300
lg_thickness = 3
img_t = (
pre_resize(
img,
target_size=target_size,
patch_size=patch_size,
)
.to(DEVICE)
.to(DTYPE)
)
img_t_lg = (
pre_resize(
img,
target_size=lg_target_size,
patch_size=lg_patch_size,
)
.to(DEVICE)
.to(DTYPE)
)
print("\nStart Outline Expansion test:")
dilate_t = dilate_cont(img_t.repeat(2, 1, 1, 1), KERNELS[thickness].to(img_t), 1)
dilate_img = Image.fromarray(to_numpy(dilate_t)[0])
dilate_img.save("./img/snow-leopard-dilate.webp", lossless=True, quality=0)
erode_t = erode_cont(img_t.repeat(2, 1, 1, 1), KERNELS[thickness].to(img_t), 1)
erode_img = Image.fromarray(to_numpy(erode_t)[0])
erode_img.save("./img/snow-leopard-erode.webp", lossless=True, quality=0)
oe_t, w = outline_expansion(
img_t.repeat(2, 1, 1, 1), thickness, thickness, patch_size, 10, 3
)
oe = Image.fromarray(to_numpy(oe_t)[0])
oe.save("./img/snow-leopard-oe.webp", lossless=True, quality=0)
w = Image.fromarray(w[0, 0].float().cpu().numpy().clip(0, 1) * 255).convert("L")
w.save("./img/snow-leopard-w.webp", lossless=True, quality=0)
print("Outline Expansion test done")
print("\nStart Pixelize test:")
print(f" Patch Size : {patch_size}")
print(f" Thickness : {thickness}")
print(f" Original Size : {img_t.shape[3]}x{img_t.shape[2]}")
print(
f" Pixelized Size: {img_t.shape[3]//patch_size}x{img_t.shape[2]//patch_size}"
)
pixel_art_t = pixelize(
img_t.repeat(2, 1, 1, 1), # for testing batch process
pixel_size=patch_size,
thickness=thickness,
do_color_match=False,
)
pixel_art = Image.fromarray(to_numpy(pixel_art_t)[0])
pixel_art.save("./img/snow-leopard-pixel.webp", lossless=True, quality=0)
print(" Pixlize test done")
pixel_art_t = pixelize(
img_t.repeat(2, 1, 1, 1),
pixel_size=patch_size,
thickness=thickness,
mode="k_centroid",
do_color_match=True,
)
pixel_art = Image.fromarray(to_numpy(pixel_art_t)[0])
pixel_art.save("./img/snow-leopard-pixel-k.webp", lossless=True, quality=0)
print(" K-Centroid test done")
pixel_art_t = pixelize(
img_t.repeat(2, 1, 1, 1),
pixel_size=patch_size,
thickness=thickness,
do_color_match=True,
do_quant=True,
num_colors=128,
dither_mode="",
)
pixel_art = Image.fromarray(to_numpy(pixel_art_t)[0])
pixel_art.save("./img/snow-leopard-pixel-128c.webp", lossless=True, quality=0)
print(" Color Quantization test done")
pixel_art_t = pixelize(
img_t.repeat(2, 1, 1, 1),
pixel_size=patch_size,
thickness=thickness,
do_color_match=True,
do_quant=True,
num_colors=128,
quant_mode="weighted-kmeans",
dither_mode="",
)
pixel_art = Image.fromarray(to_numpy(pixel_art_t)[0])
pixel_art.save(
"./img/snow-leopard-pixel-128c-weighted.webp", lossless=True, quality=0
)
print(" Weighted kmeans test done")
pixel_art_t = pixelize(
img_t.repeat(2, 1, 1, 1),
pixel_size=patch_size,
thickness=thickness,
do_color_match=True,
do_quant=True,
num_colors=256,
dither_mode="ordered",
)
pixel_art = Image.fromarray(to_numpy(pixel_art_t)[0])
pixel_art.save("./img/snow-leopard-pixel-256c-d.webp", lossless=True, quality=0)
print(" Ordered Dithering test done")
pixel_art_t = pixelize(
img_t.repeat(2, 1, 1, 1),
pixel_size=patch_size,
thickness=thickness,
do_color_match=True,
do_quant=True,
num_colors=256,
dither_mode="error_diffusion",
)
pixel_art = Image.fromarray(to_numpy(pixel_art_t)[0])
pixel_art.save("./img/snow-leopard-pixel-256c-ed.webp", lossless=True, quality=0)
print(" Error Diffusion test done")
print("\nStart Pixelize test:")
print(f" Patch Size : {lg_patch_size}")
print(f" Thickness : {lg_thickness}")
print(f" Original Size : {img_t_lg.shape[3]}x{img_t_lg.shape[2]}")
print(
f" Pixelized Size: {img_t_lg.shape[3]//lg_patch_size}x{img_t_lg.shape[2]//lg_patch_size}"
)
pixel_art_t = pixelize(
img_t_lg.repeat(2, 1, 1, 1),
pixel_size=lg_patch_size,
thickness=3,
do_color_match=True,
)
pixel_art = Image.fromarray(to_numpy(pixel_art_t)[0])
pixel_art.save("./img/snow-leopard-pixel-lg.webp", lossless=True, quality=0)
print(" Pixlize test done")
pixel_art_t = pixelize(
img_t_lg.repeat(2, 1, 1, 1),
pixel_size=lg_patch_size,
thickness=3,
do_color_match=True,
mode="k_centroid",
)
pixel_art = Image.fromarray(to_numpy(pixel_art_t)[0])
pixel_art.save("./img/snow-leopard-pixel-lg-k.webp", lossless=True, quality=0)
print(" K-Centroid test done")
pixel_art_t = pixelize(
img_t_lg.repeat(2, 1, 1, 1),
pixel_size=lg_patch_size,
thickness=3,
do_color_match=True,
do_quant=True,
num_colors=128,
dither_mode="",
)
pixel_art = Image.fromarray(to_numpy(pixel_art_t)[0])
pixel_art.save("./img/snow-leopard-pixel-lg-128c.webp", lossless=True, quality=0)
print(" Color Quantization test done")
pixel_art_t = pixelize(
img_t_lg.repeat(2, 1, 1, 1),
pixel_size=lg_patch_size,
thickness=3,
do_color_match=True,
do_quant=True,
num_colors=128,
quant_mode="weighted-kmeans",
dither_mode="",
)
pixel_art = Image.fromarray(to_numpy(pixel_art_t)[0])
pixel_art.save(
"./img/snow-leopard-pixel-lg-128c-weighted.webp", lossless=True, quality=0
)
print(" Weighted kmeans test done")
pixel_art_t = pixelize(
img_t_lg.repeat(2, 1, 1, 1),
pixel_size=lg_patch_size,
thickness=3,
do_color_match=True,
do_quant=True,
num_colors=256,
dither_mode="ordered",
)
pixel_art = Image.fromarray(to_numpy(pixel_art_t)[0])
pixel_art.save("./img/snow-leopard-pixel-lg-256c-d.webp", lossless=True, quality=0)
print(" Ordered Dithering test done")
pixel_art_t = pixelize(
img_t_lg.repeat(2, 1, 1, 1),
pixel_size=lg_patch_size,
thickness=3,
do_color_match=True,
do_quant=True,
num_colors=256,
dither_mode="error_diffusion",
)
pixel_art = Image.fromarray(to_numpy(pixel_art_t)[0])
pixel_art.save("./img/snow-leopard-pixel-lg-256c-ed.webp", lossless=True, quality=0)
print(" Error Diffusion test done")
pixeloe_env.TORCH_COMPILE = COMPILE
print("\nStart speed test:")
print(f" {target_size=}")
print(f" {patch_size=}")
print(f" {thickness=}")
print(f" {pixeloe_env.TORCH_COMPILE=}")
print(" Results:")
for bs in [1, 2, 4, 8]:
# Warmup
for _ in range(10):
pixelize(
img_t.repeat(bs, 1, 1, 1),
pixel_size=patch_size,
thickness=thickness,
do_color_match=False,
)
torch.cuda.empty_cache()
t = timeit(
"""pixelize(
img_t.repeat(bs, 1, 1, 1),
pixel_size=patch_size,
thickness=thickness,
do_color_match=False,
)""",
globals=globals(),
number=N,
)
speed = N / t * bs
print(f" bs{bs:2d}: {speed:6.3f}img/sec")
target_size = lg_target_size
patch_size = lg_patch_size
thickness = lg_thickness
print(f" {target_size=}")
print(f" {patch_size=}")
print(f" {thickness=}")
print(f" {pixeloe_env.TORCH_COMPILE=}")
print(" Results:")
for bs in [1, 2, 4, 8]:
# Warmup
for _ in range(10):
pixelize(
img_t_lg.repeat(bs, 1, 1, 1),
pixel_size=lg_patch_size,
thickness=lg_thickness,
do_color_match=False,
)
torch.cuda.empty_cache()
t = timeit(
"""pixelize(
img_t_lg.repeat(bs, 1, 1, 1),
pixel_size=lg_patch_size,
thickness=lg_thickness,
do_color_match=False,
)""",
globals=globals(),
number=N,
)
speed = N / t * bs
print(f" bs{bs:2d}: {speed:6.3f}img/sec")
print("Speed test done")