- 22 12月, 2021 1 次提交
-
-
由 Alexander Alekhin 提交于
-
- 21 12月, 2021 1 次提交
-
-
由 rogday 提交于
-
- 17 12月, 2021 3 次提交
-
-
由 Smirnov Egor 提交于
-
由 Alexander Alekhin 提交于
-
由 Maksim Shabunin 提交于
original commit: a079c2eb
-
- 15 12月, 2021 3 次提交
-
-
由 Alexander Alekhin 提交于
-
由 Smirnov Egor 提交于
fix max_unpool missing attributes, add default value of keepdims in reducemean/max/sum, add support for keepdims=true in full reduction branch, add new padding type to Pad
-
由 rogday 提交于
Onnx conformance tests * Add ONNX conformance tests * dnn(test): add filters for ONNX conformance tests * add filter lists for OCV backend * address review comments * move test_clip_inbounds to all_denylist * address clip issue * avoid empty lists Co-authored-by: NAlexander Alekhin <alexander.a.alekhin@gmail.com>
-
- 03 12月, 2021 1 次提交
-
-
由 rogday 提交于
fix ceil_mode for Average/MaxPooling * fix ceil_mode * add a comment
-
- 02 12月, 2021 1 次提交
-
-
由 Alexander Alekhin 提交于
-
- 30 11月, 2021 6 次提交
-
-
由 Smirnov Egor 提交于
-
由 Smirnov Egor 提交于
-
由 Smirnov Egor 提交于
-
由 Smirnov Egor 提交于
-
由 Andrew Ryrie 提交于
* dnn: LSTM optimisation This uses the AVX-optimised fastGEMM1T for matrix multiplications where available, instead of the standard cv::gemm. fastGEMM1T is already used by the fully-connected layer. This commit involves two minor modifications: - Use unaligned access. I don't believe this involves any performance hit in on modern CPUs (Nehalem and Bulldozer onwards) in the case where the address is actually aligned. - Allow for weight matrices where the number of columns is not a multiple of 8. I have not enabled AVX-512 as I don't have an AVX-512 CPU to test on. * Fix warning about initialisation order * Remove C++11 syntax * Fix build when AVX(2) is not available In this case the CV_TRY_X macros are defined to 0, rather than being undefined. * Minor changes as requested: - Don't check hardware support for AVX(2) when dispatch is disabled for these - Add braces * Fix out-of-bounds access in fully connected layer The old tail handling in fastGEMM1T implicitly rounded vecsize up to the next multiple of 8, and the fully connected layer implements padding up to the next multiple of 8 to cope with this. The new tail handling does not round the vecsize upwards like this but it does require that the vecsize is at least 8. To adapt to the new tail handling, the fully connected layer now rounds vecsize itself at the same time as adding the padding(which makes more sense anyway). This also means that the fully connected layer always passes a vecsize of at least 8 to fastGEMM1T, which fixes the out-of-bounds access problems. * Improve tail mask handling - Use static array for generating tail masks (as requested) - Apply tail mask to the weights as well as the input vectors to prevent spurious propagation of NaNs/Infs * Revert whitespace change * Improve readability of conditions for using AVX * dnn(lstm): minor coding style changes, replaced left aligned load
-
由 Smirnov Egor 提交于
-
- 28 11月, 2021 3 次提交
-
-
由 Alexander Alekhin 提交于
-
由 Alexander Alekhin 提交于
- remove similar test from IE scope under HAVE_INF_ENGINE
-
由 yuki takehara 提交于
resolves #21038 * remove C assert * revert C header * fix several points in review * fix test_ds.cpp
-
- 27 11月, 2021 1 次提交
-
-
由 Alexander Alekhin 提交于
-
- 12 11月, 2021 1 次提交
-
-
由 Alexander Alekhin 提交于
* dnn(ocl4dnn): fix LRN layer accuracy problems - FP16 intermediate computation is not accurate and may provide NaN values * dnn(test): update tolerance for FP16
-
- 10 11月, 2021 1 次提交
-
-
由 ZaKiiiiiiiii 提交于
fix bug: wrong output dimension when "keep_dims" is false in pooling layer. * fix bug in max layer * code align * delete permute layer and add test case * add name assert * check other cases * remove c++11 features * style:add "const" remove assert * style:sanitize file names
-
- 04 11月, 2021 1 次提交
-
-
由 Alexander Alekhin 提交于
-
- 03 11月, 2021 1 次提交
-
-
由 Alexander Alekhin 提交于
dnn(protobuf): replace deprecated calls * dnn: replace deprecated ByteSize() => ByteSizeLong() * dnn: replace deprecated calls, use GetRepeatedFieldRef
-
- 19 10月, 2021 1 次提交
-
-
由 rogday 提交于
* backport elementwise_layers refactor * keep NULL
-
- 12 10月, 2021 1 次提交
-
-
由 Smirnov Egor 提交于
-
- 11 10月, 2021 1 次提交
-
-
由 Smirnov Egor 提交于
-
- 08 10月, 2021 2 次提交
-
-
由 Alexander Alekhin 提交于
-
由 Alexander Alekhin 提交于
- prevent running code over unsupported/non-tested configurations - prevent integer div by zero
-
- 07 10月, 2021 1 次提交
-
-
由 Oliver Kuckertz 提交于
* dnn: fix unaligned memory access crash on armv7 The getTensorContent function would return a Mat pointing to some member of a Protobuf-encoded message. Protobuf does not make any alignment guarantees, which results in a crash on armv7 when loading models while bit 2 is set in /proc/cpu/alignment (or the relevant kernel feature for alignment compatibility is disabled). Any read attempt from the previously unaligned data member would send SIGBUS. As workaround, this commit makes an aligned copy via existing clone functionality in getTensorContent. The unsafe copy=false option is removed. Unfortunately, a rather crude hack in PReLUSubgraph in fact writes(!) to the Protobuf message. We limit ourselves to fixing the alignment issues in this commit, and add getTensorContentRefUnaligned to cover the write case with a safe memcpy. A FIXME marks the issue. * dnn: reduce amount of .clone() calls * dnn: update FIXME comment Co-authored-by: NAlexander Alekhin <alexander.a.alekhin@gmail.com>
-
- 06 10月, 2021 1 次提交
-
-
由 Alexander Alekhin 提交于
-
- 05 10月, 2021 1 次提交
-
-
由 Alexander Alekhin 提交于
-
- 02 10月, 2021 1 次提交
-
-
由 Alexander Alekhin 提交于
-
- 29 9月, 2021 1 次提交
-
-
由 Alexander Alekhin 提交于
-
- 17 9月, 2021 1 次提交
-
-
由 SamFC10 提交于
-
- 15 9月, 2021 1 次提交
-
-
由 rogday 提交于
Add support for YOLOv4x-mish * backport to 3.4 for supporting yolov4x-mish * add YOLOv4x-mish test * address review comments Co-authored-by: NGuo Xu <guoxu@1school.com.cn>
-
- 12 9月, 2021 1 次提交
-
-
由 Alexander Alekhin 提交于
-
- 11 9月, 2021 1 次提交
-
-
由 Zihao Mu 提交于
-
- 10 9月, 2021 2 次提交
-
-
由 Alexander Alekhin 提交于
-
由 rogday 提交于
Fix PReLU negative slope access pattern * fix prelu negative slope access pattern * change begin() to ptr()
-