dnn: added neon intrinsics implementation of fastGEMM1T function #27785
### Pull Request Readiness Checklist
See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request
- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [X] The PR is proposed to the proper branch
- This PR improves the performance of the LSTM function on ARM64 targets.
- Added a NEON intrinsics implementation of the fastGEMM1T function and enabled its use in fully connected and recurrent layers file.
- As a result, ARM64 now benefits from vectorized matrix–vector multiplications, leading to measurable performance improvements in the LSTM layer.
- This change is limited to ARM64 and does not affect other architectures.
**Performance impact:**
- The optimization significantly improves the performance of lstm functions on ARM64 targets.
<img width="930" height="313" alt="image" src="https://github.com/user-attachments/assets/92e251cd-dc6c-4cda-9586-acc19bf16dfd" />
dnn: improve performance of softmax_3d with loop unrolling #27777
### Pull Request Readiness Checklist
See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request
- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- This PR applies loop unrolling in the softmax function.
- The change does not affect functional correctness.
**Performance Improvements**
- The optimization significantly improves the performance of softmax_3d on Windows ARM64 targets.
<img width="703" height="203" alt="image" src="https://github.com/user-attachments/assets/85997c15-f543-432c-95e5-69099d71fe71" />
dnn: Tune CONV_NR_FP32 size for WASM #27773
We can see ~20% inference time reduction on local benchmark.
The local benchmark includes face detection with res10_300x300_ssd_iter_140000_fp16.caffemodel and image classification with squeezenet.onnx .
### Pull Request Readiness Checklist
See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request
- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [ ] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
- Added vector_MatShape and vector_vector_MatShape to gen_dict.json
- Implemented MatShape_to_vector_MatShape, vector_MatShape_to_MatShape, MatShape_to_vector_vector_MatShape, and vector_vector_MatShape_to_MatShape conversion functions in dnn_converters.h/cpp and Converters.java
- Added testGetLayersShapes test to verify List<List<MatShape>> conversion
- Added vector_vector_Mat to gen_dict.json
- Implemented Mat_to_vector_vector_Mat and vector_vector_Mat_to_Mat conversion functions in converters.h/cpp and Converters.java
- Added DnnForwardAndRetrieve.java test to verify List<List<Mat>> conversion : Reference: C++ test in modules/dnn/test/test_misc.cpp - TEST(Net, forwardAndRetrieve)
Close https://github.com/opencv/opencv/issues/27413
### Pull Request Readiness Checklist
See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request
- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
Feature: Add OpenVINO NPU support #27363
## Why
- OpenVINO now supports inference on integrated NPU devices in intel's Core Ultra series processors.
- Sometimes as fast as GPU, but should use considerably less power.
## How
- The NPU plugin is now available as "NPU" in openvino `ov::Core::get_available_devices()`.
- Removed the guards and checks for NPU in available targets for Inference Engine backend.
## Test example
### Pre-requisites
- Intel [Core Ultra series processor](https://www.intel.com/content/www/us/en/products/details/processors/core-ultra/edge.html#tab-blade-1-0)
- [Intel NPU driver](https://github.com/intel/linux-npu-driver/releases)
- OpenVINO 2023.3.0+ (Tested on 2025.1.0)
### Example
```cpp
#include <opencv2/dnn.hpp>
#include <iostream>
int main(){
cv::dnn::Net net = cv::dnn::readNet("../yolov8s-openvino/yolov8s.xml", "../yolov8s-openvino/yolov8s.bin");
cv::Size net_input_shape = cv::Size(640, 480);
std::cout << "Setting backend to DNN_BACKEND_INFERENCE_ENGINE and target to DNN_TARGET_NPU" << std::endl;
net.setPreferableBackend(cv::dnn::DNN_BACKEND_INFERENCE_ENGINE);
net.setPreferableTarget(cv::dnn::DNN_TARGET_NPU);
cv::Mat image(net_input_shape, CV_8UC3);
cv::randu(image, cv::Scalar(0, 0, 0), cv::Scalar(255, 255, 255));
cv::Mat blob = cv::dnn::blobFromImage(
image, 1, net_input_shape, cv::Scalar(0, 0, 0), true, false, CV_32F);
net.setInput(blob);
std::cout << "Running forward" << std::endl;
cv::Mat result = net.forward();
std::cout << "Output shape: " << result.size << std::endl; // Output shape: 1 x 84 x 6300
}
```
model files [here](https://limewire.com/d/bPgiA#BhUeSTBnMc)
docker image used to build opencv: [ghcr.io/mro47/opencv-builder](https://github.com/MRo47/opencv-builder/blob/main/Dockerfile)
Closes#26240
### Pull Request Readiness Checklist
See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request
- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
Fix#27352: Add checks before getting latest pin in Net::Impl::getLatestLayerPin() #27353
### Pull Request Readiness Checklist
Fixes#27352
See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request
- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
Fix typos #27338
### Pull Request Readiness Checklist
See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request
- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
build: fix warnings from recent gcc versions #27337
This PR addresses the following found warnings:
- [x] -Wmaybe-uninitialized
- [x] -Wunused-variable
- [x] -Wsign-compare
Tested building with GCC 14.2 (RISC-V 64).
### Pull Request Readiness Checklist
See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request
- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
TFLite fixes for Face Blendshapes V2 #27307
### Pull Request Readiness Checklist
* Scalars support
* Better handling of 1D tensors
* New ops import: SUB, SQRT, DIV, NEG, SQUARED_DIFFERENCE, SUM
* Number of NHWC<->NCHW layouts compatibility improvements
resolves#27211
**Merge with extra**: https://github.com/opencv/opencv_extra/pull/1257
See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request
- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
Explicitly specify enum type scopes to improve Java wrapper generation #27228
Changed DataLayout and ImagePaddingMode to dnn::DataLayout and dnn::ImagePaddingMode to explicitly specify their scopes. This allows gen_java.py to correctly register disc_type, preventing constructors and methods using these enum types from being skipped during Java wrapper generation.
Similarly updated QRCodeEncoder::CorrectionLevel and QRCodeEncoder::EncodeMode with explicit scope declarations.
Also added a new Java test class `DnnBlobFromImageWithParamsTest` based on: https://github.com/opencv/opencv/blob/4.x/modules/dnn/test/test_misc.cpp#L133-L243
Related issues
#23753
### Pull Request Readiness Checklist
See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request
- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake
Faster implementation of blobFromImages for cpu nchw output #26127
Faster implementation of blobFromImage and blobFromImages for
HWC cv::Mat images -> NCHW cv::Mat
case
Running time on my pc in ms:
**blobFromImage**
```
image size old new speed-up
32x32x3 0.008 0.002 4.0x
64x64x3 0.021 0.009 2.3x
128x128x3 0.164 0.037 4.4x
256x256x3 0.728 0.158 4.6x
512x512x3 3.310 0.628 5.2x
1024x1024x3 14.503 3.124 4.6x
2048x2048x3 61.647 28.049 2.2x
```
**blobFromImages**
```
image size old new speed-up
16x32x32x3 0.122 0.041 3.0x
16x64x64x3 0.790 0.165 4.8x
16x128x128x3 3.313 0.652 5.1x
16x256x256x3 13.495 3.127 4.3x
16x512x512x3 58.795 28.127 2.1x
16x1024x1024x3 251.135 121.955 2.1x
16x2048x2048x3 1023.570 487.188 2.1x
```
### Pull Request Readiness Checklist
See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request
- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
Previously, the yoloPostProcessing function assumed that the number of classes (nc) was fixed at 80. This caused incorrect behavior when a different number of classes was specified, leading to mismatched output shapes.
This update modifies the code to use the provided `nc` value dynamically, ensuring that the output shapes are correctly calculated based on the specified number of classes. This prevents issues when `nc` is not equal to 80 and allows for greater flexibility in model configurations.
An upcoming change in Protobuf will change the return types of various
methods like Descriptor::name() and Message::GetTypeName() from const
std::string& or std::string to absl::string_view. This CL fixes users
of those methods to work both before and after the change.
DNN(ONNX): Enabled several OpenCL conformance tests #26053
The tests also work in 5.x
### Pull Request Readiness Checklist
See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request
- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
Einsum buffer allocation fix#26059
This PR fixed buffer allocation issue in Einsum layer that causes segmentation fault on 32bit platforms. Related issue #26008
### Pull Request Readiness Checklist
See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request
- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [x] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
Patch to opencv_extra has the same branch name.
- [x] The feature is well documented and sample code can be built with the project CMake
Vincent Rabaud
Alexander Smorkalov
cudawarped
pratham-mcw
Jie Pan
Kavyansh Tyagi
Alexander Smorkalov
Fanchen Kong
utibenkei
fengyuentau
Kumataro
phanirithvij
Myron Rodrigues
omahs
Dmitry Kurtaev
Yuantao Feng
utibenkei
Alexander Smorkalov
Maksim Shabunin
alexlyulkov
KangJialiang
Yannis Guyon
CSBVision
Abduragim Shtanchaev