Update to use the LightGlue-ONNX fused model

[panda7281]

https://github.com/fabio-sim/LightGlue-ONNX/releases/tag/v1.0.0
Recently LightGlue-ONNX released there fused model which improve the performance substantially(~2x faster in my test) and change the format of the output tensor, making the current code not functional
output_tensors[2] is now the shape of (matchnum, 2) and the pairs are already computed(so you don't need to compute them yourself)
output_tensors[3] is now the shape of (matchnum), which denotes the matching score of each match

The following simple hack should do the work:

std::vector<int64_t> kpts0_Shape = output_tensors[0].GetTensorTypeAndShapeInfo().GetShape();
int64_t* kpts0 = (int64_t*)output_tensors[0].GetTensorMutableData<void>();
// 在Python里面是一个（batch = 1 , kpts_num , 2）的array，那么在C++里输出的长度就应该是kpts_num * 2
printf("[RESULT INFO] kpts0 Shape : (%lld , %lld , %lld)\n", kpts0_Shape[0], kpts0_Shape[1], kpts0_Shape[2]);

std::vector<int64_t> kpts1_Shape = output_tensors[1].GetTensorTypeAndShapeInfo().GetShape();
int64_t* kpts1 = (int64_t*)output_tensors[1].GetTensorMutableData<void>();
printf("[RESULT INFO] kpts1 Shape : (%lld , %lld , %lld)\n", kpts1_Shape[0], kpts1_Shape[1], kpts1_Shape[2]);

// matches: (match, [0, 1])
std::vector<int64_t> matches_Shape = output_tensors[2].GetTensorTypeAndShapeInfo().GetShape();
int64_t* matches0 = (int64_t*)output_tensors[2].GetTensorMutableData<void>();
int match_Counts = matches_Shape[0];
printf("[RESULT INFO] matches0 Shape : (%lld , %lld)\n", matches_Shape[0], matches_Shape[1]);

// match scores: (score)
std::vector<int64_t> mscore_Shape = output_tensors[3].GetTensorTypeAndShapeInfo().GetShape();
float* mscores = (float*)output_tensors[3].GetTensorMutableData<void>();

// Process kpts0 and kpts1
std::vector<cv::Point2f> kpts0_f, kpts1_f;
for (int i = 0; i < kpts0_Shape[1] * 2; i += 2)
{
    kpts0_f.emplace_back(cv::Point2f(
        (kpts0[i] + 0.5) / scales[0] - 0.5, (kpts0[i + 1] + 0.5) / scales[0] - 0.5));
}
for (int i = 0; i < kpts1_Shape[1] * 2; i += 2)
{
    kpts1_f.emplace_back(cv::Point2f(
        (kpts1[i] + 0.5) / scales[1] - 0.5, (kpts1[i + 1] + 0.5) / scales[1] - 0.5)
    );
}

std::set<std::pair<int, int> > matches;
std::vector<cv::Point2f> m_kpts0, m_kpts1;
for (int i = 0; i < matches_Shape[0] * 2; i += 2) {
    if (mscores[i/2] > cfg.threshold)
        matches.insert(std::make_pair(matches0[i], matches0[i+1]));
}

std::cout << "[RESULT INFO] matches Size : " << matches.size() << std::endl;

for (const auto& match : matches) {
    m_kpts0.emplace_back(kpts0_f[match.first]);
    m_kpts1.emplace_back(kpts1_f[match.second]);
}

keypoints_result.first = m_kpts0;
keypoints_result.second = m_kpts1;

std::cout << "[INFO] Postprocessing operation completed successfully" << std::endl;

Maybe you can add a flag for the new fused model, thanks!

[OroChippw]

This sounds very good. Since I haven’t paid attention to lightglue-related information for a while, I will upgrade and maintain the version of the code in the repository in the future.❤️

[letterso]

I have modified the project to adapt v1.0.0: Fused LightGlue-ONNX and tested it on Linux platform. Hope to help anyone who want to use LightGlue on Linux. LightGlue-OnnxRunner

[OroChippw]

Thank you for your contribution. Are you willing to initiate a pull request? After I review that there are no problems, I will merge it into the master branch and add you to the contributors.❤️

[letterso]

I don't have windows env, so I delete many codes which not be used in Linux. Avoiding to make too much change, I need some time to optimize the code before make a pull request. I will do this soon.

[wooAo]

I have modified the project to adapt v1.0.0: Fused LightGlue-ONNX and tested it on Linux platform. Hope to help anyone who want to use LightGlue on Linux. LightGlue-OnnxRunner

really help, thanks

[panda7281]

Thank you for your contribution. Are you willing to initiate a pull request? After I review that there are no problems, I will merge it into the master branch and add you to the contributors.❤️

Sorry, I'm pretty new to github and don't know how to properly submit a pull request, maybe you can merge this into your code? I'll mark this as closed
By the way, i think you should add virtual destructor to BaseFeatureMatchOnnxRunner, or the derived class won't be destruct properly

[OroChippw]

Thank you for your contribution. Are you willing to initiate a pull request? After I review that there are no problems, I will merge it into the master branch and add you to the contributors.❤️

Sorry, I'm pretty new to github and don't know how to properly submit a pull request, maybe you can merge this into your code? I'll mark this as closed By the way, i think you should add virtual destructor to BaseFeatureMatchOnnxRunner, or the derived class won't be destruct properly

Thank you for your contribution and for pointing out the bug about the destructor. I will review your code and merge it later, and update the code as soon as possible. I hope this issue can be kept open so that more people who need help can see it. ,Thanks again❤️