Voting for Voting in Online Point Cloud Object Detection

  title={Voting for Voting in Online Point Cloud Object Detection},
  author={Dominic Zeng Wang and Ingmar Posner},
  booktitle={Robotics: Science and Systems},
This paper proposes an efficient and effective scheme to applying the sliding window approach popular in computer vision to 3D data. [] Key Result For the object classes car, pedestrian and bicyclist the resulting detector achieves best-in-class detection and timing performance relative to prior art on the KITTI dataset as well as compared to another existing 3D object detection approach.

Figures and Tables from this paper

SPOT: Selective Point Cloud Voting for Better Proposal in Point Cloud Object Detection

Baselines enhanced by the Selective Point clOud voTing module can stably improve results in agreement by a large margin and achieve new stateor-the-art detection, especially under more strict evaluation metric that adopts larger IoU threshold, implying the module is the key leading to high-quality object detection in point clouds.

3D Object Detection Using Scale Invariant and Feature Reweighting Networks

A new network architecture which focuses on utilizing the front view images and frustum point clouds to generate 3D detection results and achieves better performance than the state-of-the-art methods especially when point clouds are highly sparse.

Density-Based Clustering for 3D Object Detection in Point Clouds

This work introduces a novel approach for 3D object detection that is significant in two main aspects: a cascaded modular approach that focuses the receptive field of each module on specific points in the point cloud, for improved feature learning and a class agnostic instance segmentation module that is initiated using unsupervised clustering.

Vote3Deep: Fast object detection in 3D point clouds using efficient convolutional neural networks

This paper proposes a computationally efficient approach to detecting objects natively in 3D point clouds using convolutional neural networks (CNNs) by leveraging a feature-centric voting scheme to implement novel convolutionan layers which explicitly exploit the sparsity encountered in the input.

STD: Sparse-to-Dense 3D Object Detector for Point Cloud

This work proposes a two-stage 3D object detection framework, named sparse-to-dense 3D Object Detector (STD), and implements a parallel intersection-over-union (IoU) branch to increase awareness of localization accuracy, resulting in further improved performance.

Not All Points Are Equal: Learning Highly Efficient Point-based Detectors for 3D LiDAR Point Clouds

This paper proposes a highly-efficient single-stage point-based 3D detector, termed IA-SSD, that achieves a superior speed of 80+ frames-per-second on the KITTI dataset with a single RTX2080Ti GPU.

EPN: Edge-Aware PointNet for Object Recognition from Multi-View 2.5D Point Clouds

A novel architecture coined as Edge-Aware PointNet, that incorporates geometric shape priors as binary maps, integrated in parallel with the PointNet++ framework, through convolutional neural networks (CNNs).

Recognition of Point Sets Objects in Realistic Scenes

The experimental results show that the network structure in this paper can accurately identify and classify point cloud objects in realistic scenes and maintain a certain accuracy when the number of point clouds is small, which is very robust.

PIXOR: Real-time 3D Object Detection from Point Clouds

PIXOR is proposed, a proposal-free, single-stage detector that outputs oriented 3D object estimates decoded from pixel-wise neural network predictions that surpasses other state-of-the-art methods notably in terms of Average Precision (AP), while still runs at 10 FPS.

SCNet: Subdivision Coding Network for Object Detection Based on 3D Point Cloud

A more efficient representation of 3D point clouds is proposed and SCNet, a single-stage, end-to-end 3D subdivision coding network that learns finer feature representations for vertical grids is proposed.



Towards 3D object recognition via classification of arbitrary object tracks

This paper presents a new track classification method, based on a mathematically principled method of combining log odds estimators, that is fast enough for real time use, is non-specific to object class, and performs well on the task of classifying correctly-tracked, well-segmented objects into car, pedestrian, bicyclist, and background classes.

Sliding Shapes for 3D Object Detection in Depth Images

This paper proposes to use depth maps for object detection and design a 3D detector to overcome the major difficulties for recognition, namely the variations of texture, illumination, shape, viewpoint, clutter, occlusion, self-occlusion and sensor noises.

Fast PRISM: Branch and Bound Hough Transform for Object Class Detection

This paper addresses the task of efficient object class detection by means of the Hough transform by demonstrating PRISM’s flexibility by two complementary implementations: a generatively trained Gaussian Mixture Model as well as a discriminatively trained histogram approach.

Sparse distance learning for object recognition combining RGB and depth information

This work defines a view-to-object distance where a novel view is compared simultaneously to all views of a previous object, and shows that this measure leads to superior classification performance on object category and instance recognition.

What could move? Finding cars, pedestrians and bicyclists in 3D laser data

The aim is to provide the layout of an end-to-end pipeline which, when fed by a raw stream of 3D data, produces distinct groups of points which can be fed to downstream classifiers for categorisation.

3D Object Detection and Viewpoint Estimation with a Deformable 3D Cuboid Model

This paper proposes a novel approach that extends the well-acclaimed deformable part-based model to reason in 3D, and represents an object class as a deformable 3D cuboid composed of faces and parts, which are both allowed to deform with respect to their anchors on the 3D box.

Tracking-based semi-supervised learning

It is shown experimentally that the accuracy of a classifier considered as a function of human labeling effort can be substantially improved using this method, and that a simple algorithmic speedup based on incrementally updating a boosting classifier can reduce learning time by a factor of three.

Pedestrian detection combining RGB and dense LIDAR data

A state-of-the-art deformable parts detector is trained using different configurations of optical images and their associated 3D point clouds, in conjunction and independently, leveraging upon the recently released KITTI dataset to propose novel strategies for depth upsampling and contextual fusion that together lead to detection performance which exceeds that of the RGB-only systems.

Object Detection with Discriminatively Trained Part Based Models

We describe an object detection system based on mixtures of multiscale deformable part models. Our system is able to represent highly variable object classes and achieves state-of-the-art results in

Histograms of oriented gradients for human detection

  • N. DalalB. Triggs
  • Computer Science
    2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05)
  • 2005
It is shown experimentally that grids of histograms of oriented gradient (HOG) descriptors significantly outperform existing feature sets for human detection, and the influence of each stage of the computation on performance is studied.