3D object detection combining semantic and geometric features from point clouds

Background: 3D object detection based on point clouds in road scenes has attracted much attention recently. The voxel-based methods voxelize the scene to regular grids, which can be processed with the advanced feature learning frameworks based on convolutional layers for semantic feature learning. The point-based methods can extract the geometric feature of the point due to the coordinate reservations. The combination of the two is effective for 3D object detection. However, the current methods use a voxel-based detection head with anchors for classification and localization. Although the preset anchors cover the entire scene, it is not suitable for detection tasks with larger scenes and multiple categories of objects, due to the limitation of the voxel size. Additionally, the misalignment between the predicted confidence and proposals in the Regions of the Interest (ROI) selection bring obstacles to 3D object detection. Methods: We investigate the combination of voxel-based methods and point-based methods for 3D object detection. Additionally, a voxel-to-point module that captures semantic and geometric features is proposed in the paper. The voxel-to-point module is conducive to the detection of small-size objects and avoids the presets of anchors in the inference stage. Moreover, a confidence adjustment module with the center-boundary-aware confidence attention is proposed to solve the misalignment between the predicted confidence and proposals in the regions of the interest selection. Results: The proposed method has achieved state-of-the-art results for 3D object detection in the Karlsruhe Institute of Technology and Toyota Technological Institute (KITTI) object detection dataset. Actually, as of September 19, 2021, our method ranked 1st in the 3D and Bird Eyes View (BEV) detection of cyclists tagged with difficulty level ‘easy’, and ranked 2nd in the 3D detection of cyclists tagged with ‘moderate’. Conclusions: We propose an end-to-end two-stage 3D object detector with voxel-to-point module and confidence adjustment module.

Kleinberg’s Hyper-Richness Based Fuzzy Partition Clustering for Efficient Bi-Temporal Data

Clustering is the process used for partitioning the total dataset into different classes of similar objects. The group contains knowledge about their members and also helps to understand the structure of the dataset very easily. Clustering the bitemporal data is one of the major tasks in data mining since the bitemporal datasets are very large with various attribute counts. Hence the accurate clustering is still challenging tasks. In order to improve the clustering accuracy with less complexity, Kleinberg’s Hyper-richness Bitemporal property based fuzzy c means partition Clustering (KHBP-FCMPC) technique is introduced. The KHBP-FCMPC technique partition the bitemporal dataset into number of possible groups with an improved performance rate based on a distance metric. At first, the ‘c’ numbers of clusters are initialized. The KHBP-FCMPC technique uses the core data point module and authority sector module to minimize the execution time of clustering the data points. Core data point module served as the centroid of the cluster. Each cluster contains one core data point. After that, the distance is computed with the membership function. The authority sector module assigns the data points into the cluster with minimum distance. After that, the centroid is updated and the process iterated until the convergence is met. Finally, the Kleinberg’s Hyper-richness Bitemporal property is applied to verify the total dataset equals the partition of all the data points. This property used to group the entire data points into the cluster with higher accuracy. Experimental evaluation is carried out using a temporal dataset with different factors such as clustering accuracy, false positive rate, time complexity and space complexity with a number of data points. The experimental results show that the proposed KHBP-FCMPC technique increases the bitemporal data clustering accuracy with less false positive rate, time complexity as well as space complexity. Based on the results observations, KHBP-FCMPC technique is more efficient than the state-of-the-art methods.