Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Eigen alignment issue #958

Closed
prclibo opened this issue Oct 13, 2014 · 1 comment · Fixed by #1034
Closed

Eigen alignment issue #958

prclibo opened this issue Oct 13, 2014 · 1 comment · Fixed by #1034

Comments

@prclibo
Copy link
Contributor

prclibo commented Oct 13, 2014

As stated in the Using STL Containers with Eigen in the Eigen doc, special allocator might be needed for a std::vector of Eigen::Matrix.

By a simple test grep "std::vector<Eigen" -rn | grep -v "allocator" * -rn, I noticed that some of the code does not use the allocator yet.

So just suggest to handle this issue to avoid potential problems.
@VictorLamoine
Copy link
Contributor

Thanks for the report. I'll take care of that in no more than a week.

$ grep "std::vector<Eigen" -rn | grep -vw 'aligned_allocator'
test/common/test_vector_average.cpp:48:  std::vector<Eigen::Vector3f> points;
test/common/test_vector_average.cpp:49:  std::vector<Eigen::Vector3f::Scalar> weights;
recognition/include/pcl/recognition/impl/cg/hough_3d.hpp:180:  std::vector<Eigen::Vector3d> scene_votes (n_matches);
recognition/include/pcl/recognition/impl/implicit_shape_model.hpp:144:  std::vector<Eigen::Vector3f> peaks (NUM_INIT_PTS);
recognition/include/pcl/recognition/impl/implicit_shape_model.hpp:1291:  std::vector<Eigen::Vector2f> boxes (feature_dimension);
recognition/include/pcl/recognition/impl/implicit_shape_model.hpp:1507:pcl::ism::ImplicitShapeModelEstimation<FeatureSize, PointT, NormalT>::generateRandomCenter (const std::vector<Eigen::Vector2f>& boxes,
recognition/include/pcl/recognition/implicit_shape_model.h:568:        generateRandomCenter (const std::vector<Eigen::Vector2f>& boxes, Eigen::VectorXf& center);
recognition/include/pcl/recognition/cg/hough_3d.h:450:      std::vector<Eigen::Vector3f> model_votes_;
recognition/include/pcl/recognition/face_detection/rf_face_detector_trainer.h:36:      std::vector<Eigen::Vector3f> head_center_votes_;
recognition/include/pcl/recognition/face_detection/rf_face_detector_trainer.h:37:      std::vector<std::vector<Eigen::Vector3f> > head_center_votes_clustered_;
recognition/include/pcl/recognition/face_detection/rf_face_detector_trainer.h:38:      std::vector<std::vector<Eigen::Vector3f> > head_center_original_votes_clustered_;
recognition/include/pcl/recognition/face_detection/rf_face_detector_trainer.h:39:      std::vector<Eigen::Vector3f> angle_votes_;
recognition/include/pcl/recognition/face_detection/rf_face_detector_trainer.h:41:      std::vector<Eigen::Vector3f> head_clusters_centers_;
recognition/include/pcl/recognition/face_detection/rf_face_detector_trainer.h:42:      std::vector<Eigen::Vector3f> head_clusters_rotation_;
recognition/include/pcl/recognition/face_detection/rf_face_detector_trainer.h:230:      void getDetectedFaces(std::vector<Eigen::VectorXf> & faces)
io/src/obj_io.cpp:695:  std::vector<Eigen::Vector2f> coordinates;
io/src/vtk_lib_io.cpp:370:  std::vector<Eigen::Vector2f> dummy;
geometry/include/pcl/geometry/impl/polygon_operations.hpp:183:    std::vector<Eigen::Vector3f> lines (result.size ());
common/include/pcl/common/impl/polynomial_calculations.hpp:375:      std::vector<Eigen::Matrix<real, 3, 1> >& samplePoints, unsigned int polynomial_degree, bool& error) const
common/include/pcl/common/impl/polynomial_calculations.hpp:387:      std::vector<Eigen::Matrix<real, 3, 1> >& samplePoints, unsigned int polynomial_degree,
common/include/pcl/common/impl/polynomial_calculations.hpp:418:  for (typename std::vector<Eigen::Matrix<real, 3, 1> >::const_iterator it=samplePoints.begin ();
common/include/pcl/common/impl/polynomial_calculations.hpp:468:  //for (typename std::vector<Eigen::Matrix<real, 3, 1> >::const_iterator it=samplePoints.begin ();
common/include/pcl/common/polynomial_calculations.h:96:      bivariatePolynomialApproximation (std::vector<Eigen::Matrix<real, 3, 1> >& samplePoints,
common/include/pcl/common/polynomial_calculations.h:101:      bivariatePolynomialApproximation (std::vector<Eigen::Matrix<real, 3, 1> >& samplePoints,
common/include/pcl/TextureMesh.h:101:    std::vector<std::vector<Eigen::Vector2f> >  tex_coordinates;  // UV coordinates
surface/include/pcl/surface/impl/marching_cubes_rbf.hpp:93:  std::vector<Eigen::Vector3d> centers (2*N);
surface/include/pcl/surface/impl/marching_cubes_rbf.hpp:113:        for (std::vector<Eigen::Vector3d>::const_iterator c_it = centers.begin ();
surface/include/pcl/surface/impl/marching_cubes.hpp:125:  std::vector<Eigen::Vector3f> p;
surface/include/pcl/surface/impl/gp3.hpp:141:  std::vector<Eigen::Vector2f> uvn_nn (nnn_);
surface/include/pcl/surface/impl/mls.hpp:219:    std::vector<Eigen::Vector3d> de_meaned (nn_indices.size ());
surface/include/pcl/surface/impl/texture_mapping.hpp:45:template<typename PointInT> std::vector<Eigen::Vector2f>
surface/include/pcl/surface/impl/texture_mapping.hpp:51:  std::vector<Eigen::Vector2f> tex_coordinates;
surface/include/pcl/surface/impl/texture_mapping.hpp:156:  std::vector<std::vector<Eigen::Vector2f> > texture_map;
surface/include/pcl/surface/impl/texture_mapping.hpp:161:    std::vector<Eigen::Vector2f> texture_map_tmp;
surface/include/pcl/surface/impl/texture_mapping.hpp:181:      std::vector<Eigen::Vector2f> tex_coordinates = mapTexture2Face (facet[0], facet[1], facet[2]);
surface/include/pcl/surface/impl/texture_mapping.hpp:247:  std::vector<std::vector<Eigen::Vector2f> > texture_map;
surface/include/pcl/surface/impl/texture_mapping.hpp:252:    std::vector<Eigen::Vector2f> texture_map_tmp;
surface/include/pcl/surface/impl/texture_mapping.hpp:306:  std::vector<std::vector<Eigen::Vector2f> > texture_map;
surface/include/pcl/surface/impl/texture_mapping.hpp:320:    std::vector<Eigen::Vector2f> texture_map_tmp;
surface/include/pcl/surface/impl/texture_mapping.hpp:354:  std::vector<Eigen::Vector2f> texture_map_tmp;
surface/include/pcl/surface/impl/texture_mapping.hpp:911:      std::vector<Eigen::Vector2f> dummy_container;
surface/include/pcl/surface/impl/texture_mapping.hpp:969:   std::vector<Eigen::Vector2f> dummy_container;
surface/include/pcl/surface/gp3.h:361:      std::vector<Eigen::Vector3f> coords_;
surface/include/pcl/surface/texture_mapping.h:354:      std::vector<Eigen::Vector2f>
features/include/pcl/features/principal_curvatures.h:119:      std::vector<Eigen::Vector3f> projected_normals_;
features/include/pcl/features/cvfh.h:151:      getCentroidClusters (std::vector<Eigen::Vector3f> & centroids)
features/include/pcl/features/cvfh.h:161:      getCentroidNormalClusters (std::vector<Eigen::Vector3f> & centroids)
features/include/pcl/features/cvfh.h:283:      std::vector<Eigen::Vector3f> centroids_dominant_orientations_;
features/include/pcl/features/cvfh.h:285:      std::vector<Eigen::Vector3f> dominant_normals_;
features/include/pcl/features/our_cvfh.h:195:      getCentroidClusters (std::vector<Eigen::Vector3f> & centroids)
features/include/pcl/features/our_cvfh.h:205:      getCentroidNormalClusters (std::vector<Eigen::Vector3f> & centroids)
features/include/pcl/features/our_cvfh.h:398:      std::vector<Eigen::Vector3f> centroids_dominant_orientations_;
features/include/pcl/features/our_cvfh.h:400:      std::vector<Eigen::Vector3f> dominant_normals_;
segmentation/include/pcl/segmentation/impl/crf_segmentation.hpp:223:  std::vector<Eigen::Vector3i> data;
segmentation/include/pcl/segmentation/impl/crf_segmentation.hpp:224:  std::vector<Eigen::Vector3i> color;
segmentation/include/pcl/segmentation/crf_segmentation.h:172:      std::vector<Eigen::Vector3i> data_;
segmentation/include/pcl/segmentation/crf_segmentation.h:174:      std::vector<Eigen::Vector3i> color_;
segmentation/include/pcl/segmentation/crf_segmentation.h:176:      std::vector<Eigen::Vector3f> normal_;
ml/include/pcl/ml/densecrf.h:68:      setDataVector (const std::vector<Eigen::Vector3i> data);
ml/include/pcl/ml/densecrf.h:73:      setColorVector (const std::vector<Eigen::Vector3i> color);
ml/include/pcl/ml/densecrf.h:99:      addPairwiseNormals (std::vector<Eigen::Vector3i> &coord,
ml/include/pcl/ml/densecrf.h:100:                          std::vector<Eigen::Vector3f> &normals,
ml/include/pcl/ml/densecrf.h:142:      std::vector<Eigen::Vector3i> data_;
ml/include/pcl/ml/densecrf.h:145:      std::vector<Eigen::Vector3i> color_;
ml/src/densecrf.cpp:61:pcl::DenseCrf::setDataVector (const std::vector<Eigen::Vector3i> data)
ml/src/densecrf.cpp:69:pcl::DenseCrf::setColorVector (const std::vector<Eigen::Vector3i> color)
ml/src/densecrf.cpp:136:pcl::DenseCrf::addPairwiseNormals (std::vector<Eigen::Vector3i> &coord,
ml/src/densecrf.cpp:137:                                   std::vector<Eigen::Vector3f> &normals,
filters/include/pcl/filters/impl/fast_bilateral_omp.hpp:161:    for (std::vector<Eigen::Vector2f >::iterator d = data.begin (); d != data.end (); ++d)
filters/include/pcl/filters/impl/fast_bilateral.hpp:139:    for (std::vector<Eigen::Vector2f >::iterator d = data.begin (); d != data.end (); ++d)
filters/include/pcl/filters/impl/voxel_grid_occlusion_estimation.hpp:102:                                                                std::vector<Eigen::Vector3i>& out_ray,
filters/include/pcl/filters/impl/voxel_grid_occlusion_estimation.hpp:133:pcl::VoxelGridOcclusionEstimation<PointT>::occlusionEstimationAll (std::vector<Eigen::Vector3i>& occluded_voxels)
filters/include/pcl/filters/voxel_grid.h:692:      getNeighborCentroidIndices (float x, float y, float z, const std::vector<Eigen::Vector3i> &relative_coordinates)
filters/include/pcl/filters/voxel_grid.h:697:        for (std::vector<Eigen::Vector3i>::const_iterator it = relative_coordinates.begin (); it != relative_coordinates.end (); it++)
filters/include/pcl/filters/voxel_grid_occlusion_estimation.h:112:                           std::vector<Eigen::Vector3i>& out_ray,
filters/include/pcl/filters/voxel_grid_occlusion_estimation.h:121:      occlusionEstimationAll (std::vector<Eigen::Vector3i>& occluded_voxels);
filters/include/pcl/filters/covariance_sampling.h:140:      std::vector<Eigen::Vector3f> scaled_points_;
filters/include/pcl/filters/fast_bilateral.h:125:            v_ = std::vector<Eigen::Vector2f> (width*height*depth, Eigen::Vector2f (0.0f, 0.0f));
filters/include/pcl/filters/fast_bilateral.h:167:          inline std::vector<Eigen::Vector2f >::iterator
filters/include/pcl/filters/fast_bilateral.h:171:          inline std::vector<Eigen::Vector2f >::iterator
filters/include/pcl/filters/fast_bilateral.h:175:          inline std::vector<Eigen::Vector2f >::const_iterator
filters/include/pcl/filters/fast_bilateral.h:179:          inline std::vector<Eigen::Vector2f >::const_iterator
filters/include/pcl/filters/fast_bilateral.h:184:          std::vector<Eigen::Vector2f > v_;
apps/3d_rec_framework/include/pcl/apps/3d_rec_framework/pc_source/registered_views_source.h:103:        assembleModelFromViewsAndPoses(ModelT & model, std::vector<Eigen::Matrix4f> & poses) {
apps/3d_rec_framework/include/pcl/apps/3d_rec_framework/pc_source/registered_views_source.h:162:            std::vector<Eigen::Matrix4f> poses_to_assemble_;
apps/3d_rec_framework/include/pcl/apps/3d_rec_framework/feature_wrapper/global/esf_estimator.h:28:                  std::vector<Eigen::Vector3f> & centroids)
apps/3d_rec_framework/include/pcl/apps/3d_rec_framework/feature_wrapper/global/ourcvfh_estimator.h:84:                  std::vector<Eigen::Vector3f> & centroids)
apps/3d_rec_framework/include/pcl/apps/3d_rec_framework/feature_wrapper/global/crh_estimator.h:46:                std::vector<Eigen::Vector3f> & centroids)
apps/3d_rec_framework/include/pcl/apps/3d_rec_framework/feature_wrapper/global/global_estimator.h:31:            pcl::PointCloud<FeatureT> > > & signatures, std::vector<Eigen::Vector3f> & centroids)=0;
apps/3d_rec_framework/include/pcl/apps/3d_rec_framework/feature_wrapper/global/vfh_estimator.h:31:                  std::vector<Eigen::Vector3f> & centroids)
apps/3d_rec_framework/include/pcl/apps/3d_rec_framework/feature_wrapper/global/cvfh_estimator.h:56:                std::vector<Eigen::Vector3f> & centroids)
apps/include/pcl/apps/face_detection/face_detection_apps_utils.h:112:  void displayHeads(std::vector<Eigen::VectorXf> & heads, pcl::visualization::PCLVisualizer & vis)
apps/src/render_views_tesselated_sphere.cpp:124:  std::vector<Eigen::Vector3f> cam_positions;
apps/src/face_detection/openni_face_detection.cpp:70:    std::vector<Eigen::VectorXf> heads;
apps/src/face_detection/filesystem_face_detection.cpp:85:  std::vector<Eigen::VectorXf> heads;
registration/include/pcl/registration/impl/gicp.hpp:59:                                                                                    std::vector<Eigen::Matrix3d>& cloud_covariances)
registration/include/pcl/registration/impl/gicp.hpp:76:  std::vector<Eigen::Matrix3d>::iterator matrices_iterator = cloud_covariances.begin ();
registration/include/pcl/registration/gicp.h:269:      std::vector<Eigen::Matrix3d> input_covariances_;
registration/include/pcl/registration/gicp.h:272:      std::vector<Eigen::Matrix3d> target_covariances_;
registration/include/pcl/registration/gicp.h:275:      std::vector<Eigen::Matrix3d> mahalanobis_;
registration/include/pcl/registration/gicp.h:289:                              std::vector<Eigen::Matrix3d>& cloud_covariances);
visualization/src/pcl_visualizer.cpp:3403:  std::vector<Eigen::Vector3f> cam_positions;
gpu/kinfu_large_scale/include/pcl/gpu/kinfu_large_scale/impl/standalone_marching_cubes.hpp:95:pcl::gpu::kinfuLS::StandaloneMarchingCubes<PointT>::getMeshesFromTSDFVector (const std::vector<PointCloudPtr> &tsdf_clouds, const std::vector<Eigen::Vector3f> &tsdf_offsets)
gpu/kinfu_large_scale/include/pcl/gpu/kinfu_large_scale/impl/world_model.hpp:133:pcl::kinfuLS::WorldModel<PointT>::getWorldAsCubes (const double size, std::vector<typename WorldModel<PointT>::PointCloudPtr> &cubes, std::vector<Eigen::Vector3f> &transforms, double overlap)
gpu/kinfu_large_scale/include/pcl/gpu/kinfu_large_scale/world_model.h:168:        void getWorldAsCubes (double size, std::vector<PointCloudPtr> &cubes, std::vector<Eigen::Vector3f> &transforms, double overlap = 0.0);
gpu/kinfu_large_scale/include/pcl/gpu/kinfu_large_scale/standalone_marching_cubes.h:105:      getMeshesFromTSDFVector (const std::vector<PointCloudPtr> &tsdf_clouds, const std::vector<Eigen::Vector3f> &tsdf_offsets);
gpu/kinfu_large_scale/tools/process_kinfu_large_scale_output.cpp:90:    std::vector<Eigen::Vector3f> transforms;
keypoints/src/narf_keypoint.cpp:727:  std::vector<Eigen::Matrix<RealForPolynomial, 3, 1> > sample_points;
simulation/include/pcl/simulation/range_likelihood.h:228:        std::vector<Eigen::Vector3f> vertices_;
tracking/include/pcl/tracking/impl/pyramidal_klt.hpp:485:  std::vector<Eigen::Array2f> next_pts (prev_keypoints->size ());

@VictorLamoine VictorLamoine mentioned this issue Dec 6, 2014
Merged
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

Eigen with std::vector fixes
2 participants
@prclibo @VictorLamoine