[PDF] RIU-Net: Embarrassingly simple semantic segmentation of 3D LiDAR point cloud | Semantic Scholar (2024)

[PDF] Semantic Reader

This work proposes an end-to-end architecture for LiDAR point cloud semantic segmentation that efficiently solves the problem as an image processing problem and outperforms the state-of-the-art by a large margin on the KITTI dataset.

An end-to-end architecture for LiDAR point cloud semantic segmentation that efficiently solves the problem as an image processing problem and outperforms the state-of-the-art by a large margin on the KITTI dataset.

Experiments show that RangeSeg outperforms the state-of-the-art semantic segmentation methods with enormous speedup and improves the instance-level segmentation performance on small and far objects.

This paper proposes a lightweight semantic segmentation network for large-scale point clouds which consists of grid subsampling, dilated convolutions, and Gaussian error linear unit activation for gaining better performance.

A compact but flexible framework is introduced for performing segmentation to the semantics of point clouds concurrently, contribute to more precise segmentation and based on semantics labels, a novel DCRF model is elaborated to refine the result of segmentation.

A review of existing 3D datasets and 3D semantic segmentation methods and efforts to solve data hungry problems are summarized for both 3D LiDAR-focused methods and general-purpose methods.

This article aims to provide a comprehensive survey of 3D point cloud and DL-based methods for scene understanding in autonomous driving, mainly divided into two subtasks: object detection and semantic segmentation.

This is the first work to analyze the data hunger problem for 3D semantic segmentation using deep learning techniques that are addressed in the literature review, statistical analysis, andCross-dataset and cross-algorithm experiments.

Experiments show that the multi-level feature fusion method can increase the performance of road segmentation, and the FCN model has a general segmentation effect in complex road scenes.

It is demonstrated that non-stationary objects can be detected using neural network models trained with 2-D grayscale images in the supervised or unsupervised training process, which makes it possible to alleviate the lack of large datasets of 3-D laser scans with point-wise annotations for non- stationary objects.

...

...

This paper takes the spherical image, which is transformed from the 3D LiDAR point clouds, as input of the convolutional neural networks (CNNs) to predict the point-wise semantic map, which makes it quite compatible for autonomous driving applications.

This work promotes an alternative approach by using this image representation of the 3D point cloud, taking advantage of the fact that the problem of disocclusion has been intensively studied in the 2D image processing community over the past decade.

An end-to-end pipeline called SqueezeSeg based on convolutional neural networks (CNN), which takes a transformed LiDAR point cloud as input and directly outputs a point-wise label map, which is then refined by a conditional random field (CRF) implemented as a recurrent layer.

It is argued that the organization of 3D point clouds can be efficiently captured by a structure called superpoint graph (SPG), derived from a partition of the scanned scene into geometrically hom*ogeneous elements.

This work introduces a new model SqueezeSegV2, which is more robust against dropout noises in LiDAR point cloud and therefore achieves significant accuracy improvement, and a domain-adaptation training pipeline consisting of three major components: learned intensity rendering, geodesic correlation alignment, and progressive domain calibration.

This paper designs a novel type of neural network that directly consumes point clouds, which well respects the permutation invariance of points in the input and provides a unified architecture for applications ranging from object classification, part segmentation, to scene semantic parsing.

VoxelNet is proposed, a generic 3D detection network that unifies feature extraction and bounding box prediction into a single stage, end-to-end trainable deep network and learns an effective discriminative representation of objects with various geometries, leading to encouraging results in3D detection of pedestrians and cyclists.

A LIDAR-based perception system for ground robot mobility, consisting of 3D object detection, classification and tracking, enabling it to safely navigate in urban traffic-like scenarios as well as in off-road convoy scenarios.

The proposed algorithm can reliably detect all major planes in the scene at a frame rate of more than 35Hz for 640×480 point clouds, which to the best of the knowledge is much faster than state-of-the-art algorithms.

...

...