Created release v19.17
dlib/CMakeLists.txt | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
Make cusolver finding work for really old versions of cmake with CUDA 10.1
dlib/CMakeLists.txt | 10 ++++++++++ 1 file changed, 10 insertions(+)
Automatically test for NEON instructions when building python extensions and enable them if available.
dlib/cmake_utils/set_compiler_specific_options.cmake | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)
Added more dianostic messages from cmake.
dlib/CMakeLists.txt | 3 +++ 1 file changed, 3 insertions(+)
Fixed more gcc 8 warnings.
dlib/cpp_tokenizer/cpp_tokenizer_kernel_1.h | 2 +- dlib/svm/pegasos.h | 1 + 2 files changed, 2 insertions(+), 1 deletion(-)
Fixed compiler warnings in gcc 8
dlib/array/array_kernel.h | 4 ++-- dlib/array2d/array2d_kernel.h | 4 ++-- dlib/array2d/serialize_pixel_overloads.h | 20 ++++++++++---------- .../binary_search_tree/binary_search_tree_kernel_1.h | 2 +- .../binary_search_tree/binary_search_tree_kernel_2.h | 2 +- dlib/gui_widgets/fonts.cpp | 4 ++-- dlib/hash_map/hash_map_kernel_1.h | 2 +- dlib/hash_set/hash_set_kernel_1.h | 2 +- dlib/hash_table/hash_table_kernel_1.h | 2 +- dlib/hash_table/hash_table_kernel_2.h | 2 +- dlib/map/map_kernel_1.h | 2 +- dlib/optimization/optimization_bobyqa.h | 2 +- dlib/queue/queue_kernel_1.h | 2 +- dlib/queue/queue_kernel_2.h | 2 +- dlib/sequence/sequence_kernel_1.h | 2 +- dlib/sequence/sequence_kernel_2.h | 2 +- dlib/set/set_kernel_1.h | 2 +- dlib/sliding_buffer/sliding_buffer_kernel_1.h | 2 +- dlib/stack/stack_kernel_1.h | 2 +- dlib/static_set/static_set_kernel_1.h | 2 +- dlib/test/compress_stream.cpp | 2 +- 21 files changed, 33 insertions(+), 33 deletions(-)
Changed to avoid compiler warning.
examples/dnn_semantic_segmentation_train_ex.cpp | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
Work around bug in nvcc in cuda 10.1
dlib/matrix/matrix.h | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-)
Made cusolver finding work in newer versions of cmake and cuda
dlib/CMakeLists.txt | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-)
Add python api that generates desciptor(s) from the aligned image(s) (#1667)
python_examples/face_recognition.py | 17 +++++++++ tools/python/src/face_recognition.cpp | 72 ++++++++++++++++++++++++++++++++++- 2 files changed, 87 insertions(+), 2 deletions(-)
Make sure CUDA_VERSION is defined (#1663)
dlib/cuda/gpu_data.cpp | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)
Exposing chinese_whispers directly, closes #1642 (#1644)
tools/python/src/face_recognition.cpp | 35 +++++++++++++++++++ tools/python/test/test_chinese_whispers.py | 56 ++++++++++++++++++++++++++++++ 2 files changed, 91 insertions(+)
fixed check for excessive detections in loss_mmod_ (#1625)
dlib/dnn/loss.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
Remove Python 3.5 tests on AppVeyor because something is wrong with python 3.5 on appveyor and the other tests are sufficient for python coverage.
dlib/appveyor/python.yml | 2 -- 1 file changed, 2 deletions(-)
Apply the #1514 fix even on non-Windows platforms, and change all remaining cudaStreamSynchronize calls (#1596)
dlib/cuda/gpu_data.cpp | 51 ++++++++++++++++++++++++-------------------------- 1 file changed, 24 insertions(+), 27 deletions(-)
Reliably detect AVX2 at runtime on Linux (#1614)
dlib/simd/simd_check.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
merged
Fixed broken memcpy overload.
dlib/cuda/cuda_data_ptr.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
Add U-net style skip connections to the semantic-segmentation example (#1600)
dlib/dnn/layers.h | 100 +++++++++++++++++ dlib/dnn/layers_abstract.h | 50 +++++++++ dlib/test/dnn.cpp | 4 +- examples/dnn_semantic_segmentation_ex.cpp | 8 +- examples/dnn_semantic_segmentation_ex.h | 137 ++++++++++++++++-------- examples/dnn_semantic_segmentation_train_ex.cpp | 29 ++--- 6 files changed, 262 insertions(+), 66 deletions(-)
remove register keyword from jdhuff for C++17 (#1612)
dlib/external/libjpeg/jdhuff.cpp | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
Add pointwise_divide function (#1586)
dlib/dnn/layers.h | 2 +- dlib/matrix/matrix_utilities.h | 153 ++++++++++++++++++++++++++++++++ dlib/matrix/matrix_utilities_abstract.h | 40 +++++++++ dlib/test/matrix.cpp | 4 + dlib/test/matrix2.cpp | 34 ++++--- 5 files changed, 221 insertions(+), 12 deletions(-)
Fixed code not compiling.
dlib/cuda/cuda_data_ptr.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
Update CmakeLists: add png for arm (#1580)
dlib/CMakeLists.txt | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
Add padding argument support for Python compute_face_descriptor methods (#1578)
python_examples/face_recognition.py | 6 +++++- tools/python/src/face_recognition.cpp | 33 +++++++++++++++++++++------------ 2 files changed, 26 insertions(+), 13 deletions(-)
printing error (#1568)
python_examples/train_object_detector.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
Add MKL with TBB support to CMake (#1561)
dlib/CMakeLists.txt | 12 ++++++++++- dlib/cmake_utils/find_blas.cmake | 45 ++++++++++++++++++++++++++++++---------- 2 files changed, 45 insertions(+), 12 deletions(-)
Minor changes to avoid bugs in some compilers.
tools/python/src/decision_functions.cpp | 2 +- tools/python/src/face_recognition.cpp | 10 +++++----- tools/python/src/global_optimization.cpp | 14 +++++++------- tools/python/src/gui.cpp | 2 +- tools/python/src/image2.cpp | 4 ++-- tools/python/src/image4.cpp | 2 +- tools/python/src/image_dataset_metadata.cpp | 8 ++++---- tools/python/src/numpy_returns.cpp | 4 ++-- tools/python/src/shape_predictor.cpp | 2 +- 9 files changed, 24 insertions(+), 24 deletions(-)
Fix find_blas.cmake bug: find mkl fft on windows. (#1557)
dlib/cmake_utils/find_blas.cmake | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+)
Set CMake policy CMP0048 to NEW when available. (#1549)
dlib/CMakeLists.txt | 5 +++++ 1 file changed, 5 insertions(+)
merged
When building python extensions in Visual Studio, never link to any system copies of libpng or libjpeg, instead, use the copies what come with dlib. We are doing this because too many Visual Studio users have busted copies of these libraries on their machines. This will make things just always work.
dlib/CMakeLists.txt | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-)
Apparently fix #1513 by avoiding the cudaStreamSynchronize call (#1514)
dlib/cuda/gpu_data.cpp | 24 ++++++++++++++++++++++++ 1 file changed, 24 insertions(+)
Fixed random_forest_regression.h warnings (#1544)
dlib/random_forest/random_forest_regression.h | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-)
Avoid compiler bug
tools/python/src/object_detection.cpp | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
Added weighted labels to loss_binary_log layer. (#1538)
dlib/dnn/loss.h | 10 +++++----- dlib/dnn/loss_abstract.h | 14 ++++++++++---- 2 files changed, 15 insertions(+), 9 deletions(-)
Add header needed in some contexts
dlib/sockets/sockets_kernel_2.cpp | 1 + 1 file changed, 1 insertion(+)
Added --box-images to imglab
tools/imglab/src/main.cpp | 31 +++++++++++++++++++++++++++++-- 1 file changed, 29 insertions(+), 2 deletions(-)
Record last changeset and set PATCH version to 99
dlib/CMakeLists.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
Created release v19.16
dlib/CMakeLists.txt | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)