Open Source Computer Vision Library https://opencv.org/

You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

Tag: Branch: Tree: e1ce2146f5

2.4

3.4

4.x

5.x

master

next

2.2

2.3.0

2.3.1

2.4.0

2.4.1

2.4.10

2.4.10.1

2.4.10.2

2.4.10.3

2.4.10.4

2.4.11

2.4.12

2.4.12.1

2.4.12.2

2.4.12.3

2.4.13

2.4.13.1

2.4.13.2

2.4.13.3

2.4.13.4

2.4.13.5

2.4.13.6

2.4.13.7

2.4.2

2.4.3

2.4.3-rc

2.4.3.1

2.4.3.2

2.4.4

2.4.4-beta

2.4.5

2.4.6

2.4.6.1

2.4.6.2

2.4.6.2-rc1

2.4.6.2r2

2.4.6.2r3

2.4.7

2.4.7-rc1

2.4.7.1

2.4.7.2

2.4.8

2.4.8.1

2.4.8.2

2.4.8.3

2.4.9

2.4.9.1

3.0-ocl-tech-preview

3.0-ocl-tp2

3.0.0

3.0.0-alpha

3.0.0-beta

3.0.0-rc1

3.1.0

3.2.0

3.2.0-rc

3.3.0

3.3.0-cvsdk

3.3.0-rc

3.3.1

3.3.1-cvsdk

3.4.0

3.4.0-rc

3.4.1

3.4.1-cvsdk

3.4.10

3.4.11

3.4.12

3.4.13

3.4.14

3.4.15

3.4.16

3.4.17

3.4.18

3.4.19

3.4.2

3.4.2-openvino

3.4.20

3.4.3

3.4.3-openvino

3.4.4

3.4.5

3.4.6

3.4.7

3.4.8

3.4.9

4.0.0

4.0.0-alpha

4.0.0-beta

4.0.0-openvino

4.0.0-rc

4.0.1

4.0.1-openvino

4.1.0

4.1.0-openvino

4.1.1

4.1.1-openvino

4.1.2

4.1.2-openvino

4.10.0

4.10.0-kleidicv

4.2.0

4.2.0-openvino

4.3.0

4.3.0-openvino

4.3.0-openvino-2020.3.0

4.4.0

4.4.0-openvino

4.5.0

4.5.0-openvino

4.5.1

4.5.1-openvino

4.5.2

4.5.2-openvino

4.5.3

4.5.3-openvino

4.5.3-openvino-2021.4.1

4.5.3-openvino-2021.4.2

4.5.4

4.5.5

4.5.5-openvino-2022.1.0

4.6.0

4.7.0

4.8.0

4.8.1

4.9.0

Branches Tags

${ item.name }

Create tag ${ searchTerm }

Create branch ${ searchTerm }

from 'e1ce2146f5'

${ noResults }

opencv/modules/dnn/perf

History

Andrew Ryrie ea7d4be3f8 Merge pull request #20658 from smbz:lstm_optimisation ... * 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		3 years ago
..
perf_caffe.cpp	test: fix usage of findDataFile()	6 years ago
perf_common.cpp	cmake: fix build of dnn tests with shared common code	6 years ago
perf_convolution.cpp	dnn(perf): update convolution tests	3 years ago
perf_convolution1d.cpp	Merge pull request #18783 from sl-sergei:fix_conv1d	4 years ago
perf_convolution3d.cpp	Merge pull request #18783 from sl-sergei:fix_conv1d	4 years ago
perf_layer.cpp	dnn: fix OpenCL implementation of Slice layer	4 years ago
perf_main.cpp	dnn: move module from opencv_contrib	8 years ago
perf_net.cpp	dnn(test): update tests for OpenVINO 2021.2	4 years ago
perf_precomp.hpp	dnn(perf): fix and merge Convolution tests	6 years ago
perf_recurrent.cpp	Merge pull request #20658 from smbz:lstm_optimisation	3 years ago