Cant use yolov10 in ONNX format with opencv

C++
,
1

I’ve trained yolov10 model and exported it in .onnx. Then I want to use it in my c++ code, but getting error during processing node with 2 inputs and 2 outputs.
Here is full output with this error:

[ERROR:0@0.126] global onnx_importer.cpp:1040 cv::dnn::dnn4_v20251223::ONNXImporter::handleNode DNN/ONNX: ERROR during processing node with 2 inputs and 2 outputs: [TopK]:(onnx_node!/model.23/TopK_1) from domain='ai.onnx'
OpenCV: terminate handler is called! The last OpenCV error is:
OpenCV(4.13.0) Error: Unspecified error (> Node [TopK@ai.onnx]:(onnx_node!/model.23/TopK_1) parse error: OpenCV(4.13.0) C:\Users\pankovn\Downloads\opencv-4.13.0\modules\dnn\src\layers\topk_layer.cpp:102: error: (-2:Unspecified error) in function 'bool __cdecl cv::dnn::TopKLayerImpl::getMemoryShapes(const class std::vector<class std::vector<int,class std::allocator<int> >,class std::allocator<class std::vector<int,class std::allocator<int> > > > &,const int,class std::vector<class std::vector<int,class std::allocator<int> >,class std::allocator<class std::vector<int,class std::allocator<int> > > > &,class std::vector<class std::vector<int,class std::allocator<int> >,class std::allocator<class std::vector<int,class std::allocator<int> > > > &) const'
> > TopK: K is out of range (expected: 'K < input_shape[axis_normalized]'), where
> >     'K' is 300
> > must be less than
> >     'input_shape[axis_normalized]' is 300
> ) in cv::dnn::dnn4_v20251223::ONNXImporter::handleNode

OpenCV version: 4.13.0 Operating System / Platform: Windows 11

2

top K… out of K

looks like a situation that might not necessarily be an error, but might be pointless (a no-op).

you could browse the (open and closed) issues on opencv’s github. if there is nothing, you could add this. make sure to present easy steps to reproduce. this may require the weights of the model. if any officially provided weights do the job of reproing the issue, that’d be convenient.