SuperPoint Rpautrat (MIT license) (#1603)

* initial python implementation of superpoint rpautrat

* working python implementation of superpoint

* small tweaks to try and work around the GIL issue

* fix missing os import

* begging cpp impl of superpoint python model

* finishing cpp superpoint impl using the same SPDetector interface

* finalizing superpoint cpp impl, working with cpu but needs to be cleaned

* fixing feature matching by reworking nms and filtering logic

* speeding up nms with batched operations and cleaning up

* updating conversion script

* adding args for image dimensions and cuda usage to model tracer

* wiring up UI to parse superpoint params

* adding label to superpoint rpautrat ui

* reverting some unintended ui changes

* typo

* oneline revert

* removing nms and threshold filtering from cpp, this is handled internally by the superpoint model

* using python interface to run the superpoint _to_torchscript.py script at runtime

* generate and load model file on the first incoming frame

* reverting unintented change

* ui changes appear mysteriously again, reverting

* remove topk from model to prevent issue when number of detected keypoints is lower then the k value (scripting fails in this case)

* cleaning up for review

* change dest for model file and remove debug logs

* bump patch and add version comment

* rm unucessary comment

* use resources to load file to ensure it works when rtabmap isn't built from source

* execute with pybind runpy instead of system call

* only build superpoint rpautrat if we have torch and python support

* changing param to accept a path to the weights .pth file directly

* parse the default working directory to save the model file in

* rm extraneous change

* rm setters and re-initialize the detector whenever params change. We cannot support changing params after the model is constructed

* update UI text

* only enable superpoint rpautrat when built with python and torch

* more build information regarding superpoint rpautrat

* generate temporary python script in the working dir

* rm unecessary changes to rtabmap_superpoint.py

* fix comment and only add repo root to sys.path

* introduce a new parameter for the superpoint python model definition

* execute script from string instead of writing to a file

* wrap parse params in a single compiler directive

* remove descriptor spatial matching logic and rely on upstream RTAB-Map processes to take the top-K desc and kpts

* only add python script to resources if built with superpoint rpautrat support. supress warning with type casting

* isolate pybind11 setup so it can't affect any other modules or potential regnerations of the model file

* update about dialog to show superpoint rpautrat

* remove debug logs

---------

Co-authored-by: Felix Toft <felix@robust.ai>
This commit is contained in:
Felix Toft
2025-11-07 17:22:35 -08:00
committed by GitHub
co-authored by Felix Toft
parent f44a4fc478
commit 5e4fd171e2
16 changed files with 2083 additions and 1178 deletions
+30 -1
View File
@@ -104,6 +104,7 @@ namespace rtabmap {
class ORBextractor;
class SPDetector;
class SPDetectorRpautrat;
class Stereo;
#if CV_MAJOR_VERSION < 3
@@ -129,7 +130,8 @@ public:
kFeatureSurfFreak=12, //new 0.20.4
kFeatureGfttDaisy=13, //new 0.20.6
kFeatureSurfDaisy=14, //new 0.20.6
kFeaturePyDetector=15}; //new 0.20.8
kFeaturePyDetector=15, //new 0.20.8
kFeatureSuperPointRpautrat=16}; // new 0.23.3
static std::string typeName(Type type)
{
@@ -164,6 +166,8 @@ public:
return "GFTT+Daisy";
case kFeatureSurfDaisy:
return "SURF+Daisy";
case kFeatureSuperPointRpautrat:
return "SUPERPOINT-RPAUTRAT";
default:
return "Unknown";
}
@@ -626,6 +630,31 @@ private:
bool cuda_;
};
//SuperPointRpautrat
class RTABMAP_CORE_EXPORT SuperPointRpautrat : public Feature2D
{
public:
SuperPointRpautrat(const ParametersMap & parameters = ParametersMap());
virtual ~SuperPointRpautrat();
virtual void parseParameters(const ParametersMap & parameters);
virtual Feature2D::Type getType() const { return kFeatureSuperPointRpautrat; }
private:
virtual std::vector<cv::KeyPoint> generateKeypointsImpl(const cv::Mat & image, const cv::Rect & roi, const cv::Mat & mask = cv::Mat());
virtual cv::Mat generateDescriptorsImpl(const cv::Mat & image, std::vector<cv::KeyPoint> & keypoints) const;
cv::Ptr<SPDetectorRpautrat> superPoint_;
std::string superpointWeightsPath_;
std::string superpointModelPath_;
std::string outputDir_;
float threshold_;
bool nms_;
int minDistance_;
bool cuda_;
};
//GFTT_DAISY
class RTABMAP_CORE_EXPORT GFTT_DAISY : public GFTT
{
+11 -4
View File
@@ -252,9 +252,9 @@ class RTABMAP_CORE_EXPORT Parameters
RTABMAP_PARAM(Kp, NndrRatio, float, 0.8, "NNDR ratio (A matching pair is detected, if its distance is closer than X times the distance of the second nearest neighbor.)");
#if CV_MAJOR_VERSION > 2 && !defined(HAVE_OPENCV_XFEATURES2D)
// OpenCV>2 without xFeatures2D module doesn't have BRIEF
RTABMAP_PARAM(Kp, DetectorStrategy, int, 8, "0=SURF 1=SIFT 2=ORB 3=FAST/FREAK 4=FAST/BRIEF 5=GFTT/FREAK 6=GFTT/BRIEF 7=BRISK 8=GFTT/ORB 9=KAZE 10=ORB-OCTREE 11=SuperPoint 12=SURF/FREAK 13=GFTT/DAISY 14=SURF/DAISY 15=PyDetector");
RTABMAP_PARAM(Kp, DetectorStrategy, int, 8, "0=SURF 1=SIFT 2=ORB 3=FAST/FREAK 4=FAST/BRIEF 5=GFTT/FREAK 6=GFTT/BRIEF 7=BRISK 8=GFTT/ORB 9=KAZE 10=ORB-OCTREE 11=SuperPoint 12=SURF/FREAK 13=GFTT/DAISY 14=SURF/DAISY 15=PyDetector 16=SuperPoint-Rpautrat");
#else
RTABMAP_PARAM(Kp, DetectorStrategy, int, 6, "0=SURF 1=SIFT 2=ORB 3=FAST/FREAK 4=FAST/BRIEF 5=GFTT/FREAK 6=GFTT/BRIEF 7=BRISK 8=GFTT/ORB 9=KAZE 10=ORB-OCTREE 11=SuperPoint 12=SURF/FREAK 13=GFTT/DAISY 14=SURF/DAISY 15=PyDetector");
RTABMAP_PARAM(Kp, DetectorStrategy, int, 6, "0=SURF 1=SIFT 2=ORB 3=FAST/FREAK 4=FAST/BRIEF 5=GFTT/FREAK 6=GFTT/BRIEF 7=BRISK 8=GFTT/ORB 9=KAZE 10=ORB-OCTREE 11=SuperPoint 12=SURF/FREAK 13=GFTT/DAISY 14=SURF/DAISY 15=PyDetector 16=SuperPoint-Rpautrat");
#endif
RTABMAP_PARAM(Kp, TfIdfLikelihoodUsed, bool, true, "Use of the td-idf strategy to compute the likelihood.");
RTABMAP_PARAM(Kp, Parallelized, bool, true, "If the dictionary update and signature creation were parallelized.");
@@ -346,6 +346,13 @@ class RTABMAP_CORE_EXPORT Parameters
RTABMAP_PARAM(SuperPoint, NMSRadius, int, 4, uFormat("[%s=true] Minimum distance (pixels) between keypoints.", kSuperPointNMS().c_str()));
RTABMAP_PARAM(SuperPoint, Cuda, bool, true, "Use Cuda device for Torch, otherwise CPU device is used by default.");
RTABMAP_PARAM_STR(SuperPointRpautrat, WeightsPath, "", "[Required] SuperPoint weights file (*.pth).");
RTABMAP_PARAM_STR(SuperPointRpautrat, ModelPath, "", "[Required] SuperPoint python model file (superpoint_pytorch.py).");
RTABMAP_PARAM(SuperPointRpautrat, Threshold, float, 0.005, "Detector response threshold to accept keypoint.");
RTABMAP_PARAM(SuperPointRpautrat, NMS, bool, true, "If true, non-maximum suppression is applied to detected keypoints.");
RTABMAP_PARAM(SuperPointRpautrat, NMSRadius, int, 4, uFormat("[%s=true] Minimum distance (pixels) between keypoints.", kSuperPointRpautratNMS().c_str()));
RTABMAP_PARAM(SuperPointRpautrat, Cuda, bool, true, "Use Cuda device for Torch, otherwise CPU device is used by default.");
RTABMAP_PARAM_STR(PyDetector, Path, "", "Path to python script file (see available ones in rtabmap/corelib/src/python/*). See the header to see where the script should be copied.");
RTABMAP_PARAM(PyDetector, Cuda, bool, true, "Use cuda.");
@@ -704,9 +711,9 @@ class RTABMAP_CORE_EXPORT Parameters
RTABMAP_PARAM(Vis, Iterations, int, 300, "Maximum iterations to compute the transform.");
#if CV_MAJOR_VERSION > 2 && !defined(HAVE_OPENCV_XFEATURES2D)
// OpenCV>2 without xFeatures2D module doesn't have BRIEF
RTABMAP_PARAM(Vis, FeatureType, int, 8, "0=SURF 1=SIFT 2=ORB 3=FAST/FREAK 4=FAST/BRIEF 5=GFTT/FREAK 6=GFTT/BRIEF 7=BRISK 8=GFTT/ORB 9=KAZE 10=ORB-OCTREE 11=SuperPoint 12=SURF/FREAK 13=GFTT/DAISY 14=SURF/DAISY 15=PyDetector");
RTABMAP_PARAM(Vis, FeatureType, int, 8, "0=SURF 1=SIFT 2=ORB 3=FAST/FREAK 4=FAST/BRIEF 5=GFTT/FREAK 6=GFTT/BRIEF 7=BRISK 8=GFTT/ORB 9=KAZE 10=ORB-OCTREE 11=SuperPoint 12=SURF/FREAK 13=GFTT/DAISY 14=SURF/DAISY 15=PyDetector 16=SuperPoint-Rpautrat");
#else
RTABMAP_PARAM(Vis, FeatureType, int, 6, "0=SURF 1=SIFT 2=ORB 3=FAST/FREAK 4=FAST/BRIEF 5=GFTT/FREAK 6=GFTT/BRIEF 7=BRISK 8=GFTT/ORB 9=KAZE 10=ORB-OCTREE 11=SuperPoint 12=SURF/FREAK 13=GFTT/DAISY 14=SURF/DAISY 15=PyDetector");
RTABMAP_PARAM(Vis, FeatureType, int, 6, "0=SURF 1=SIFT 2=ORB 3=FAST/FREAK 4=FAST/BRIEF 5=GFTT/FREAK 6=GFTT/BRIEF 7=BRISK 8=GFTT/ORB 9=KAZE 10=ORB-OCTREE 11=SuperPoint 12=SURF/FREAK 13=GFTT/DAISY 14=SURF/DAISY 15=PyDetector 16=SuperPoint-Rpautrat");
#endif
RTABMAP_PARAM(Vis, MaxFeatures, int, 1000, "0 no limits.");
RTABMAP_PARAM(Vis, SSC, bool, false, "If true, SSC (Suppression via Square Covering) is applied to limit keypoints.");