1698c72 | Víctor Ponce-López | 15 November 2018, 19:06:43 UTC | Added the undeclared negligible penalisation also for non-semantic control case | 15 November 2018, 19:06:43 UTC |
13d2e89 | Víctor Ponce-López | 15 November 2018, 19:03:22 UTC | Fixed Penalisation within the Discriminator Loss | 15 November 2018, 19:03:22 UTC |
4023928 | Víctor Ponce-López | 25 October 2018, 17:39:17 UTC | New Penalisation and Resizing method | 25 October 2018, 17:39:17 UTC |
dc9d04c | Víctor Ponce-López | 25 October 2018, 15:48:54 UTC | Discriminator Loss Penalisation for real images semantically constrained by OpenFace | 25 October 2018, 15:48:54 UTC |
56b27c8 | Víctor Ponce-López | 25 October 2018, 15:22:57 UTC | Update README.md | 25 October 2018, 15:22:57 UTC |
8c6be7d | Víctor Ponce-López | 13 April 2018, 19:00:09 UTC | Avoid type error in tensorflow 1.5.0 | 13 April 2018, 19:00:09 UTC |
09d9adc | Víctor Ponce-López | 19 March 2018, 22:29:34 UTC | Update dcgan.yml | 19 March 2018, 22:29:34 UTC |
d9ec3d5 | Víctor Ponce-López | 19 March 2018, 22:24:47 UTC | Update README.md | 19 March 2018, 22:24:47 UTC |
2b17800 | Víctor Ponce-López | 19 March 2018, 18:23:30 UTC | Merge pull request #1 from vponcelo/vponcelo-temporal Added frame gaps functionality to generate from training temporal data sequences | 19 March 2018, 18:23:30 UTC |
29a7416 | Víctor Ponce-López | 19 March 2018, 18:20:32 UTC | Update README.md | 19 March 2018, 18:20:32 UTC |
bb53b27 | Víctor Ponce-López | 19 March 2018, 18:14:28 UTC | modelling with frame gaps for temporal sequences Modification of the model for training images within the specified gap over the consecutive frames. | 19 March 2018, 18:14:28 UTC |
8b56e6a | Víctor Ponce-López | 19 March 2018, 18:06:45 UTC | Added frame gaps parameter Train within a gap of specified seconds on the frame rate to skip redundant images in the temporal dimension | 19 March 2018, 18:06:45 UTC |
3532685 | Zhedong Zheng | 16 October 2017, 11:41:44 UTC | Update utils.py | 16 October 2017, 11:41:44 UTC |
07ccf59 | Zhedong Zheng | 16 October 2017, 11:40:43 UTC | Update README.md | 16 October 2017, 11:40:43 UTC |
d3b6ab4 | Zhedong Zheng | 16 October 2017, 07:01:30 UTC | Update README.md | 16 October 2017, 07:01:30 UTC |
849519b | Zhedong Zheng | 16 October 2017, 05:47:46 UTC | Update README.md | 16 October 2017, 05:47:46 UTC |
bc236fb | Layumi | 16 October 2017, 05:00:14 UTC | remove web | 16 October 2017, 05:00:14 UTC |
c1fb9ee | Zhedong Zheng | 16 October 2017, 04:54:50 UTC | Delete mnist3.png | 16 October 2017, 04:54:50 UTC |
440d7c4 | Zhedong Zheng | 16 October 2017, 04:54:41 UTC | Delete test_2016-01-27 15:07:47.png | 16 October 2017, 04:54:41 UTC |
f1f6ea6 | Zhedong Zheng | 16 October 2017, 04:54:35 UTC | Delete test_2016-01-27 15:09:00.png | 16 October 2017, 04:54:35 UTC |
688392d | Zhedong Zheng | 16 October 2017, 04:54:27 UTC | Delete test_2016-01-27 15:09:50.png | 16 October 2017, 04:54:27 UTC |
1e5f5ad | Zhedong Zheng | 16 October 2017, 04:54:20 UTC | Delete training.gif | 16 October 2017, 04:54:20 UTC |
2bab932 | Zhedong Zheng | 16 October 2017, 04:54:13 UTC | Delete test_2016-01-27 15:09:46.png | 16 October 2017, 04:54:13 UTC |
947ec5d | Zhedong Zheng | 16 October 2017, 04:54:06 UTC | Delete test_2016-01-27 15:08:45.png | 16 October 2017, 04:54:06 UTC |
f1e03ad | Zhedong Zheng | 16 October 2017, 04:53:58 UTC | Delete test_2016-01-27 15:08:57.png | 16 October 2017, 04:53:58 UTC |
d5b2d67 | Zhedong Zheng | 16 October 2017, 04:53:51 UTC | Delete test_2016-01-27 15:08:54.png | 16 October 2017, 04:53:51 UTC |
185bd78 | Zhedong Zheng | 16 October 2017, 04:53:45 UTC | Delete d_loss.png | 16 October 2017, 04:53:45 UTC |
b29505c | Zhedong Zheng | 16 October 2017, 04:53:36 UTC | Delete d_hist.png | 16 October 2017, 04:53:36 UTC |
8a34d5c | Zhedong Zheng | 16 October 2017, 04:53:29 UTC | Delete d__hist.png | 16 October 2017, 04:53:29 UTC |
7805221 | Zhedong Zheng | 16 October 2017, 04:53:21 UTC | Delete custom_dataset.png | 16 October 2017, 04:53:21 UTC |
704a813 | Zhedong Zheng | 16 October 2017, 04:50:37 UTC | Delete training.mp4 | 16 October 2017, 04:50:37 UTC |
022dc7f | Zhedong Zheng | 16 October 2017, 04:50:30 UTC | Delete random.mp4 | 16 October 2017, 04:50:30 UTC |
d445694 | Zhedong Zheng | 16 October 2017, 04:50:22 UTC | Delete background.mp4 | 16 October 2017, 04:50:22 UTC |
0bf1b1a | Zhedong Zheng | 16 October 2017, 03:29:26 UTC | Update README.md | 16 October 2017, 03:29:26 UTC |
e12ba48 | Zhedong Zheng | 11 October 2017, 12:46:31 UTC | Update README.md | 11 October 2017, 12:46:31 UTC |
5a3cd90 | Zhedong Zheng | 11 October 2017, 12:43:00 UTC | Update README.md | 11 October 2017, 12:43:00 UTC |
5ca9448 | Zhedong Zheng | 11 October 2017, 12:41:27 UTC | Update README.md | 11 October 2017, 12:41:27 UTC |
be523dc | Zhedong Zheng | 11 October 2017, 12:40:33 UTC | Update main.py | 11 October 2017, 12:40:33 UTC |
72bd7b6 | Zhedong Zheng | 11 October 2017, 12:39:55 UTC | Update utils.py | 11 October 2017, 12:39:55 UTC |
390147f | Zhedong Zheng | 11 October 2017, 12:21:46 UTC | Create dcgan.yml | 11 October 2017, 12:21:46 UTC |
5ff4c21 | Zhedong Zheng | 11 October 2017, 12:20:27 UTC | Update README.md | 11 October 2017, 12:20:27 UTC |
3939671 | Zhedong Zheng | 11 October 2017, 12:18:54 UTC | Update main.py | 11 October 2017, 12:18:54 UTC |
b138300 | Taehoon Kim | 20 July 2017, 16:21:01 UTC | Merge pull request #199 from ngc92/unify_code_paths combined code paths for with/without y_dim | 20 July 2017, 16:21:01 UTC |
1c5c3d5 | ngc92 | 13 July 2017, 19:33:10 UTC | combined code paths for with/without y_dim | 13 July 2017, 19:33:10 UTC |
a8bab57 | Taehoon Kim | 22 June 2017, 00:08:00 UTC | Merge pull request #192 from minhwanoh/patch-1 Update utils.py | 22 June 2017, 00:08:00 UTC |
358cdf1 | Min-hwan Oh | 20 June 2017, 20:21:43 UTC | Update utils.py Fixed manifold dimension bug | 20 June 2017, 20:21:43 UTC |
cb63afa | Taehoon Kim | 19 June 2017, 10:25:36 UTC | Merge pull request #191 from ngc92/fix_non_square_sample_counts fix generation of sample image non square number of images | 19 June 2017, 10:25:36 UTC |
15c3c19 | Taehoon Kim | 19 June 2017, 10:24:01 UTC | Merge pull request #189 from richardjdavies/patch-2 Remove unused placeholder variable in build_model | 19 June 2017, 10:24:01 UTC |
dc74d44 | Taehoon Kim | 19 June 2017, 10:23:14 UTC | Merge pull request #188 from richardjdavies/patch-1 Fix layer 4 name in discriminator | 19 June 2017, 10:23:14 UTC |
b963154 | ngc92 | 18 June 2017, 20:41:12 UTC | fix generation of sample image non square number of images | 18 June 2017, 20:41:12 UTC |
32a8d9b | Richard Davies | 09 June 2017, 14:30:10 UTC | Remove unused placeholder variable in build_model | 09 June 2017, 14:30:10 UTC |
4165cab | Richard Davies | 08 June 2017, 13:07:11 UTC | Fix layer 4 name in discriminator | 08 June 2017, 13:07:11 UTC |
432665f | Taehoon Kim | 01 June 2017, 00:10:57 UTC | Merge pull request #173 from Titinious/patch-1 Bug Fix for Checking Grayscale Image | 01 June 2017, 00:10:57 UTC |
c66492c | Zackary P. T. Sin | 20 May 2017, 12:25:22 UTC | Bug Fix for Checking Grayscale Image I have attempted to use the program with grayscale images. c_dim will become n for a n x n image (e.g. c_dim = 96 for a 96 x 96 image). By modifying as such, grayscale image could be successfully detected and c_dim will be appropriately assigned with 1. | 20 May 2017, 12:25:22 UTC |
afdb010 | Taehoon Kim | 26 April 2017, 08:30:14 UTC | refactor codes and fix bug of c_dim | 26 April 2017, 08:30:14 UTC |
1fdb04b | Taehoon Kim | 26 April 2017, 07:50:23 UTC | fix error of c_dim | 26 April 2017, 07:50:23 UTC |
7e97809 | Taehoon Kim | 26 April 2017, 06:18:12 UTC | get rid of c_dim. fix #162 | 26 April 2017, 06:18:12 UTC |
7f7175c | Taehoon Kim | 14 April 2017, 22:33:41 UTC | add related works | 14 April 2017, 22:33:41 UTC |
07abd2b | Taehoon Kim | 06 April 2017, 08:42:52 UTC | argument error. fix #158 | 06 April 2017, 08:42:52 UTC |
9f8f6a6 | Taehoon Kim | 04 April 2017, 06:02:56 UTC | Merge pull request #156 from b-liu14/patch-2 Deel with ValueError in utils.imsave when using mnist. | 04 April 2017, 06:02:56 UTC |
254a9ac | Taehoon Kim | 04 April 2017, 06:00:35 UTC | Merge pull request #157 from carpedm20/revert-155-patch-1 Revert "Use visualize option." | 04 April 2017, 06:00:35 UTC |
bd9ef42 | Taehoon Kim | 04 April 2017, 06:00:25 UTC | Revert "Use visualize option." | 04 April 2017, 06:00:25 UTC |
af8a901 | Taehoon Kim | 04 April 2017, 05:59:38 UTC | Merge pull request #155 from b-liu14/patch-1 Use visualize option. | 04 April 2017, 05:59:38 UTC |
5354dde | 刘斌 | 04 April 2017, 03:29:48 UTC | Deel with ValueError: 'arr' does not have a suitable array shape for any mode. When using mnist dataset and c_dim equals 1, the shape of value of merge(images, size) is (224, 224, 1), while the function scipy.misc.imsave just support these three type of shape: MxN or MxNx3 or MxNx4(https://docs.scipy.org/doc/scipy/reference/generated/scipy.misc.imsave.html). So we need to squeeze it first. | 04 April 2017, 03:29:48 UTC |
e08bf5d | 刘斌 | 04 April 2017, 03:24:25 UTC | Use visualize option. | 04 April 2017, 03:24:25 UTC |
8b2e226 | Taehoon Kim | 03 April 2017, 11:46:20 UTC | Merge branch 'master' of github.com:carpedm20/DCGAN-tensorflow | 03 April 2017, 11:46:20 UTC |
7d853b1 | Taehoon Kim | 03 April 2017, 11:02:45 UTC | update CelebA url | 03 April 2017, 11:46:08 UTC |
aa10b2f | Taehoon Kim | 03 April 2017, 11:01:17 UTC | Merge pull request #152 from viligabi/master fix: imsave() raised error by images with 1 channels | 03 April 2017, 11:01:17 UTC |
10d379e | viligabi | 01 April 2017, 17:35:48 UTC | fix: imsave() raised exeption by images with 1 channels scipy.misc.imsave can handle only the images with the following dims: HxW, HxWx3, HxWx4 imsave() function now handles only images with channels 1, 3 or 4 | 01 April 2017, 17:35:48 UTC |
0188e8e | Taehoon Kim | 01 April 2017, 09:08:14 UTC | Merge pull request #151 from viligabi/save_images()_fix fix: save_images() filename issue | 01 April 2017, 09:08:14 UTC |
5f66ae2 | Taehoon Kim | 01 April 2017, 09:08:00 UTC | Merge pull request #150 from viligabi/merge_fix fix: merge() now can handle images with other than 3 channels | 01 April 2017, 09:08:00 UTC |
3134375 | Taehoon Kim | 01 April 2017, 09:07:36 UTC | Merge pull request #149 from viligabi/master fix: Issue #135 | 01 April 2017, 09:07:36 UTC |
bf8a199 | Taehoon Kim | 01 April 2017, 09:07:17 UTC | Merge pull request #148 from crizCraig/patch-1 Fix dataset download command | 01 April 2017, 09:07:17 UTC |
98970f7 | viligabi | 01 April 2017, 09:03:29 UTC | fix: save_images() filename issue | 01 April 2017, 09:03:29 UTC |
50505f3 | viligabi | 01 April 2017, 08:47:10 UTC | fix: merge() now can handle images with other than 3 channels | 01 April 2017, 08:47:10 UTC |
d9ce252 | viligabi | 01 April 2017, 08:24:18 UTC | fix: save_images() function was called with fixed 8x8 manifold size -> by other than 64 batch size the images were not generated | 01 April 2017, 08:24:18 UTC |
c4db4a6 | Craig Quiter | 30 March 2017, 19:37:01 UTC | Fix dataset download command | 30 March 2017, 19:37:01 UTC |
6e74fd6 | Taehoon Kim | 24 March 2017, 08:41:48 UTC | Merge pull request #144 from XDXX/patch-1 fixed invalid markdown image link | 24 March 2017, 08:41:48 UTC |
ea7a2a1 | 郝智翔 | 24 March 2017, 08:30:11 UTC | fixed invalid markdown image link The current link 'assets/test_2016-01-27 15:08:54.png' has a space so it's invalid in markdown and cannot show the image. | 24 March 2017, 08:30:11 UTC |
ca2edf6 | Taehoon Kim | 19 March 2017, 15:39:50 UTC | Merge pull request #140 from johnhany/master Fix ValueError when batch_size is not by default | 19 March 2017, 15:39:50 UTC |
158b77d | John Hany | 19 March 2017, 08:25:58 UTC | Fix ValueError when batch_size is not by default Add `sample_num=FLAGS.batch_size` to DCGAN initializing. Resolving issue #135 | 19 March 2017, 08:25:58 UTC |
3faddba | Taehoon Kim | 12 March 2017, 01:22:05 UTC | add show_variables and add g_bn0 is_train=False | 12 March 2017, 01:22:05 UTC |
934fe3d | Taehoon Kim | 12 March 2017, 00:44:02 UTC | Merge pull request #134 from viligabi/master Fix: tensorboard scalar graph is overlapping with previous data when training is started from checkpoint | 12 March 2017, 00:44:02 UTC |
01ce6cf | viligabi | 11 March 2017, 10:39:36 UTC | Fix: tensorboard scalar graph is overlapping with previous data when training is started from checkpoint In case the training is loaded from checkpoint, the new training results will be overlapped by the old ones since the counter is not reinitalized by loading. Solution: load() is now returning with the counter when the checkpoint was saved. In case the load is successful the counter will be reinitialized | 11 March 2017, 10:39:36 UTC |
be49b99 | Taehoon Kim | 01 March 2017, 02:13:48 UTC | height -> width typo, Fix #126 | 01 March 2017, 02:13:48 UTC |
a49ea7e | Taehoon Kim | 23 February 2017, 14:57:12 UTC | Merge pull request #122 from asispatra/patch-1 visualize() failed for batch_size more than 64 | 23 February 2017, 14:57:12 UTC |
128f002 | Asis Patra | 23 February 2017, 04:47:31 UTC | visualize() failed for batch_size more than 64 The code was always saving the images with in a 8x8 image frame. So it can hold maximum 64 images. That's why it was not failing if batch_size is less or equals to 64. If batch_size is more than 64, it was giving error. As a solution we can increase the image frame dimensions as square. So that the batch size does not overflow in the image frame. | 23 February 2017, 04:47:31 UTC |
33ece0f | Taehoon Kim | 22 February 2017, 16:14:44 UTC | Merge pull request #118 from asispatra/patch-1 sigmoid_cross_entropy_with_logits function was returning None | 22 February 2017, 16:14:44 UTC |
4ce73aa | Asis Patra | 22 February 2017, 09:20:43 UTC | sigmoid_cross_entropy_with_logits function was returning None In the sigmoid_cross_entropy_with_logits function definition it was not returning anything. It should return something which can be used by tf.reduce_mean(). | 22 February 2017, 09:20:43 UTC |
3d7395a | Taehoon Kim | 20 February 2017, 18:22:00 UTC | sigmoid_cross_entropy_with_logits kawrgs and int problem. fix #109, #114 | 20 February 2017, 18:22:04 UTC |
9166ef9 | Taehoon Kim | 08 February 2017, 11:29:27 UTC | tf.concat_v2 removed, fix #99 | 08 February 2017, 11:29:27 UTC |
66884d3 | Taehoon Kim | 08 February 2017, 09:53:19 UTC | fix concat error. fix #98 | 08 February 2017, 10:09:41 UTC |
e4b395b | Taehoon Kim | 08 February 2017, 08:16:45 UTC | Merge pull request #97 from Randl/patch-1 Fix python 3 compatibility | 08 February 2017, 08:16:45 UTC |
cbd36ba | Evgeniy Zheltonozhskiy | 08 February 2017, 07:18:24 UTC | Fix python 3 compatibility `xrange` should be imported | 08 February 2017, 07:18:24 UTC |
74d32ae | Taehoon Kim | 08 February 2017, 00:12:05 UTC | Merge pull request #96 from Randl/Randl-patch-2 Add concat to support TF 1.0 | 08 February 2017, 00:12:05 UTC |
ceab058 | Taehoon Kim | 08 February 2017, 00:11:09 UTC | Merge pull request #95 from Randl/Randl-patch-1 Correct calculation of `s_h2`, etc. | 08 February 2017, 00:11:09 UTC |
84745a0 | Evgeniy Zheltonozhskiy | 07 February 2017, 18:07:22 UTC | Add concat to support TF 1.0 | 07 February 2017, 18:07:22 UTC |
0e603b1 | Evgeniy Zheltonozhskiy | 07 February 2017, 18:02:56 UTC | Fix `concat_v2` to support TF 1.0 | 07 February 2017, 18:02:56 UTC |
48115cf | Evgeniy Zheltonozhskiy | 07 February 2017, 17:44:36 UTC | import math | 07 February 2017, 17:44:36 UTC |
e51b5bf | Evgeniy Zheltonozhskiy | 07 February 2017, 17:42:33 UTC | Right size calculation in one more place | 07 February 2017, 17:42:33 UTC |