Commit a451e792 authored by Davis King's avatar Davis King

updated docs

parent efd94561
...@@ -23,8 +23,8 @@ New Features and Improvements: ...@@ -23,8 +23,8 @@ New Features and Improvements:
- Made log loss layers more numerically stable. - Made log loss layers more numerically stable.
- Upgraded the con layer so you can set the number of rows or columns to - Upgraded the con layer so you can set the number of rows or columns to
0 in the layer specification. Doing this means "make the filter cover 0 in the layer specification. Doing this means "make the filter cover
the whole input image dimension". So it's an easy way to make a filter the whole input image dimension". This provides an easy way to make a
sized exactly so that it will have one output along that dimension, filter sized so it will have one output along that dimension,
effectively making it like a fully connected layer operating on a row effectively making it like a fully connected layer operating on a row
or column. or column.
- Added support for non-scale-invariant MMOD. - Added support for non-scale-invariant MMOD.
...@@ -48,9 +48,9 @@ New Features and Improvements: ...@@ -48,9 +48,9 @@ New Features and Improvements:
- Added disjoint_subsets_sized - Added disjoint_subsets_sized
- Added Python APIs: get_face_chips(), count_steps_without_decrease(), - Added Python APIs: get_face_chips(), count_steps_without_decrease(),
count_steps_without_decrease_robust(), and jitter_image(). count_steps_without_decrease_robust(), and jitter_image().
- Various improvements to cmake scripts: e.g. improved warning and error - Various improvements to CMake scripts: e.g. improved warning and error
messages, added USE_NEON_INSTRUCTIONS option. messages, added USE_NEON_INSTRUCTIONS option.
- chol() will use a banded cholesky algorithm for banded matrices, making it - chol() will use a banded Cholesky algorithm for banded matrices, making it
much faster in these cases. much faster in these cases.
- Changed the timing code to use the C++11 high resolution clock and - Changed the timing code to use the C++11 high resolution clock and
atomics. This makes the timing code a lot more precise. atomics. This makes the timing code a lot more precise.
...@@ -78,7 +78,6 @@ Non-Backwards Compatible Changes: ...@@ -78,7 +78,6 @@ Non-Backwards Compatible Changes:
this change reverts back to the previous behavior. this change reverts back to the previous behavior.
Bug fixes: Bug fixes:
- Fixed DLIB_ISO_CPP_ONLY not building.
- Fixed toMat() not compiling in some cases. - Fixed toMat() not compiling in some cases.
- Significantly reduced the compile time of the DNN example programs in - Significantly reduced the compile time of the DNN example programs in
visual studio. visual studio.
...@@ -110,7 +109,8 @@ Bug fixes: ...@@ -110,7 +109,8 @@ Bug fixes:
some of these problems, in particular, there were some cases where some of these problems, in particular, there were some cases where
unloading dlib's python extension would deadlock. This should now be unloading dlib's python extension would deadlock. This should now be
fixed. fixed.
- Fixed DLIB_STACK_TRACE macro not compiling. - Fixed compile time errors when either of these macros were enabled:
DLIB_STACK_TRACE, DLIB_ISO_CPP_ONLY.
</current> </current>
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment