Davis King
04e034a70f
Made object part annotations shown on the image_display scale relative to the size of their bounding box.
2017-09-09 09:23:24 -04:00
Davis King
85c94dc1e4
removed extraneous print statement.
2017-09-07 19:11:18 -04:00
Davis King
ef57efc742
Fix compiler error in visual studio.
2017-09-06 10:33:53 -04:00
Davis King
16327b658e
Made input_rgb_image_pyramid use the GPU if available. So it's now a lot
...
faster. The CPU path is also a lot faster as well.
2017-09-06 07:55:27 -04:00
Davis King
abf392538e
Made cpu version of tt::resize_bilinear() use SIMD instructions and multiple
...
CPU cores.
2017-09-06 07:53:49 -04:00
Davis King
07483b1fe9
Added overloads of sub_image() that take raw pointers so you can make
...
sub_images of anything. Also made the resize_image() functions a little more
flexible and compatible with the other image interpolation code in dlib.
2017-09-05 22:06:07 -04:00
Davis King
a0febe7f69
Clarified spec
2017-09-05 06:16:32 -04:00
Davis King
9e54a2088a
Refactored create_tiled_pyramid() code to remove unnecessary copying. It's now
...
about 2x faster. The code is cleaner too.
2017-09-04 21:56:24 -04:00
Davis King
8c2d87db09
Added find_pyramid_down_output_image_size()
2017-09-04 16:27:56 -04:00
Davis King
9ba4b45ffd
Upgraded resize_bilinear() to let the user specify independent row and channel
...
stride values. This lets you run the tensor resizing routine on subwindows in
a tensor.
2017-09-04 15:42:34 -04:00
Davis King
f5a68ded86
mmod_options would pick bad window sizes in some corner cases. This has been fixed.
2017-09-03 17:16:56 -04:00
Davis King
14ae224ecd
Made it so you can press END or i to ignore boxes in imglab. I did this
...
because it's a much less stressing hand motion to hit END that i in most cases.
2017-09-03 12:54:18 -04:00
Davis King
38db737860
Fixed a bug in test_object_detection_function() which was just introduced. The
...
problem was that it wasn't counting detection boxes as false alarms if the
detection boxes had labels that never appeared in the truth dataset.
2017-09-03 12:15:36 -04:00
Davis King
770781ac1e
Added a label field to mmod_rect and updated code that uses this object to use
...
the label in the way you would expect. In particular, loss_mmod_ is now a
multi-class loss and therefore capable of learning a detector that can output
detections with different labels.
2017-09-02 22:07:07 -04:00
Davis King
b56f73cef2
Added box_percent_covered()
2017-09-02 08:42:36 -04:00
Davis King
ac5206eb10
merged
2017-09-01 23:08:25 -04:00
Davis King
e5df3d8e49
merged
2017-09-01 23:07:12 -04:00
Davis King
60d474a8b7
merged
2017-09-01 21:24:38 -04:00
Davis King
972bf3c46b
Made the box colors in imglab spread more evenly over the color space.
2017-09-01 17:38:41 -04:00
Davis King
511c05adbe
Updated spec
2017-09-01 17:35:32 -04:00
Davis King
5940ebf375
Made it so crossed out overlay boxes are drawn slightly faded.
2017-09-01 16:47:27 -04:00
Davis King
f53c6c32fc
Fixed a bug in the extract layer that trigged when a tensor with a different
...
number of samples than the tensor used to initialize the network was passed
through the layer.
2017-08-31 09:27:04 -04:00
Davis King
9e67724d43
Clarified spec
2017-08-31 09:25:27 -04:00
Davis King
b8dc0044fb
Improved error message.
2017-08-31 09:25:07 -04:00
Davis King
0bb3d9f1d1
Fixed error in input_rgb_image_pyramid::image_contained_point(). The function
...
didn't properly account for the new padding on the edge of the pyramid and
might erroneously indicate that a point wasn't inside the original image when
really it was.
2017-08-31 07:21:07 -04:00
Davis King
7b9154ff92
Record last changeset and set PATCH version to 99
2017-08-28 18:18:45 -04:00
Davis King
e62f9ee5b3
Added tag v19.6 for changeset 3eaa0e35b1b4
2017-08-28 18:17:46 -04:00
Davis King
cf0f61921c
Created release v19.6
2017-08-28 18:17:46 -04:00
Davis King
ae9fe86daa
updated docs
2017-08-28 18:17:40 -04:00
Davis King
7767f68038
merged
2017-08-28 08:00:43 -04:00
Evgeniy Fominov
2a33dc2ac6
fix static runtime in cuda with clang and windows ( #773 )
2017-08-28 08:00:17 -04:00
Davis King
ed9199ecbf
saving more pypi notes
2017-08-27 20:04:31 -04:00
Davis King
50e8b23f0a
Record last changeset and set PATCH version to 99
2017-08-27 19:35:22 -04:00
Davis King
330a53dca6
Added tag v19.5 for changeset 9121e039950d
2017-08-27 19:34:24 -04:00
Davis King
763a6ef850
Created release v19.5
2017-08-27 19:34:24 -04:00
Davis King
a4eb894ff0
Set this file to executable
2017-08-27 19:13:46 -04:00
Davis King
3263488114
merged
2017-08-27 18:41:27 -04:00
Guillaume Ramé
bbf3d987b2
improvements to cnn face detection python interface ( #780 )
...
* improvements to cnn face detection interface
* mmod rectangle object renaming. possibility to set batch size in multi image detection. Added check to make sure images are all the same size.
2017-08-27 18:40:31 -04:00
Davis King
32e00f1aea
updated docs
2017-08-27 18:40:21 -04:00
Davis King
71db541a92
Minor cleanup
2017-08-27 18:18:00 -04:00
Davis King
306fd3af10
updated docs
2017-08-27 13:03:58 -04:00
Davis King
3211da440d
Yet more comments
2017-08-27 12:01:24 -04:00
Davis King
a362305e1b
cleanup
2017-08-27 10:46:49 -04:00
Davis King
5e9f094d5a
Fixed compiler warning
2017-08-27 09:23:57 -04:00
Davis King
71f78c1ee6
Fixed compiler warnings
2017-08-27 09:23:49 -04:00
Davis King
88c1bf3b46
Added youtube links
2017-08-27 09:17:05 -04:00
Davis King
efb1d83dd3
More comments
2017-08-27 09:11:49 -04:00
Davis King
54de7a8546
Added more comments
2017-08-27 08:29:36 -04:00
Davis King
aec695fc26
Cleanup
2017-08-27 07:31:50 -04:00
Davis King
f6e23cabc7
Gave upsample_image_dataset() an option to limit upsampling on really large images.
2017-08-27 07:31:38 -04:00
