forked from chuanli11/CNNMRF
-
Notifications
You must be signed in to change notification settings - Fork 0
/
run_syn.lua
executable file
·49 lines (41 loc) · 4.63 KB
/
run_syn.lua
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
require 'paths'
paths.dofile('mylib/helper.lua')
-----------------------------------------
-- Parameters:
-----------------------------------------
-- content_name: the content image located in folder "data/content". Notice for free synthesis this image is only used for initialization (and only when "ini_method" is set to "image")
-- style_name: the style image located in folder "data/style"
-- ini_method: initial method, set to "image" to use the content image as the initialization; set to "random" to use random noise (of the same size as the content image).
-- max_size: maximum size of the synthesis image. Default value 384. Larger image needs more time and memory.
-- scaler: relative expansion from example to result. Default value 2.
-- num_res: number of resolutions. Default value 3. Notice the lowest resolution image should be larger than the patch size otherwise it won't synthesize.
-- num_iter: number of iterations for each resolution. Default value 100 for all resolutions.
-- mrf_layers: the layers for MRF constraint. Usualy layer 21 alone already gives decent results. Including layer 12 may improve the results but at significantly more computational cost.
-- mrf_weight: weight for each MRF layer. Default value 1e-4. For free texture synthesis it can be seen as the "learning rate" in gradient decent.
-- mrf_patch_size: the patch size for MRF constraint. Default value 3. This value is defined seperately for each MRF layer.
-- mrf_num_rotation: To matching objects of different poses. Default value 0. This value is shared by all MRF layers. The total number of rotatoinal copies is "2 * mrf_num_rotation + 1"
-- mrf_num_scale: To matching objects of different scales. Default value 0. This value is shared by all MRF layers. The total number of scaled copies is "2 * mrf_num_scale + 1"
-- mrf_sample_stride: stride to sample mrf on style image. Default value 2. This value is defined seperately for each MRF layer.
-- mrf_synthesis_stride: stride to sample mrf on synthesis image. Default value 2. This value is defined seperately for each MRF layer.
-- mrf_confidence_threshold: threshold for filtering out bad matching. Default value 0 -- means we keep all matchings. This value is defined seperately for all layers.
-- tv_weight: TV smoothness weight. Default value 1e-3.
-- mode: speed or memory. Try 'speed' if you have a GPU with more than 4GB memory, and try 'memory' otherwise. The 'speed' mode is significantly faster (especially for synthesizing high resolutions) at the cost of higher GPU memory.
-- gpu_chunck_size_1: Size of chunks to split feature maps along the channel dimension. This is to save memory when normalizing the matching score in mrf layers. Use large value if you have large gpu memory. As reference we use 256 for Titan X, and 32 for Geforce GT750M 2G.
-- gpu_chunck_size_2: Size of chuncks to split feature maps along the y dimension. This is to save memory when normalizing the matching score in mrf layers. Use large value if you have large gpu memory. As reference we use 16 for Titan X, and 2 for Geforce GT750M 2G.
-- backend: Use 'cudnn' for CUDA-enabled GPUs or 'clnn' for OpenCL.
-----------------------------------------
-- Reference tests
-----------------------------------------
-- speed mode V.S. memory mode (Titan X 12G)
-- {'2', '2', 'random', 384, 2, 3, {100, 100, 100}, {12, 21}, {1e-4, 1e-4}, {3, 3}, 1, 1, {2, 2}, {2, 2}, {0, 0}, 1e-3, 'speed', 256, 16, 'cudnn'}, -- 131 seconds
-- {'2', '2', 'random', 384, 2, 3, {100, 100, 100}, {12, 21}, {1e-4, 1e-4}, {3, 3}, 1, 1, {2, 2}, {2, 2}, {0, 0}, 1e-3, 'memory', 256, 16, 'cudnn'} -- 172 seconds
-- speed mode V.S. memory mode (Geforce GT750M 2G)
-- {'2', '2', 'random', 384, 2, 3, {100, 100, 100}, {12, 21}, {1e-4, 1e-4}, {3, 3}, 1, 1, {2, 2}, {2, 2}, {0, 0}, 1e-3, 'speed', 256, 16, 'cudnn'}, -- 552 seconds (gpu streching, not recommended)
-- {'2', '2', 'random', 384, 2, 3, {100, 100, 100}, {12, 21}, {1e-4, 1e-4}, {3, 3}, 1, 1, {2, 2}, {2, 2}, {0, 0}, 1e-3, 'memory', 256, 16, 'cudnn'}, -- 1506 seconds
-- speed mode V.S. memory mode (Sapphire Radeon R9 280 3G)
-- {'2', '2', 'random', 384, 2, 3, {100, 100, 100}, {12, 21}, {1e-4, 1e-4}, {3, 3}, 1, 1, {2, 2}, {2, 2}, {0, 0}, 1e-3, 'speed', 256, 16, 'clnn'}, -- 193 seconds (240 seconds total)
-- {'2', '2', 'random', 384, 2, 3, {100, 100, 100}, {12, 21}, {1e-4, 1e-4}, {3, 3}, 1, 1, {2, 2}, {2, 2}, {0, 0}, 1e-3, 'memory', 256, 16, 'clnn'}, -- 175 seconds (216 seconds total)
local list_params = {
{'2', '2', 'random', 384, 2, 3, {100, 100, 100}, {12, 21}, {1e-4, 1e-4}, {3, 3}, 1, 1, {2, 2}, {2, 2}, {0, 0}, 1e-3, 'memory', 256, 16, 'cudnn'},
}
run_tests(require 'syn_CNNMRF_wrapper', list_params)