Vehicle Detection from 3D Lidar Using Fully Convolutional Network

By Bo Li, Tianlei Zhang, Tian Xia. 2016

655 citations recorded by the source corpus.

Convolutional network techniques have recently achieved great success in vision based detection tasks. This paper introduces the recent development of our research on transplanting the fully convolutional network technique to the detection tasks on 3D range scan data. Specifically, the scenario is set as the vehicle detection task from the range data of Velodyne 64E lidar. We proposes to present the data in a 2D point map and use a single 2D end-to-end fully convolutional network to predict the objectness confidence and the bounding boxes simultaneously. By carefully design the bounding box encoding, it is able to predict full 3D bounding boxes even using a 2D convolutional network. Experiments on the KITTI dataset shows the state-of-the-art performance of the proposed method.

Subfields: Automotive Engineering, Computer Vision and Pattern Recognition, Environmental Engineering, Instrumentation

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Automotive Engineering

In Computer Vision and Pattern Recognition

In Environmental Engineering

In Instrumentation

Citations in the corpus

VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection (2017, methodology): “Several methods project point clouds into a perspective view and apply image-based feature extraction techniques [30, 15, 22].”
A Survey of Deep Network Solutions for Learning Control in Robotics: From Reinforcement to Imitation (2016, neutral): “n Training Precision Model Source Test Efﬁciency Lenz et al. (2015b) RGBD DNN Benchmark X Object Detection Grasping FF F Redmon and Angelova (2015) RGBD CNN Benchmark X Object Detection Grasping F FF Li et al. (2016) 3D Lidar CNN Benchmark Object Detection Auto Drive F F Maturana and Scherer (2015a)”
Static-Dynamic Co-Teaching for Class-Incremental 3D Object Detection (2021, background): “Many deep learning-based approaches (Li, Zhang, and Xia 2016; Chen et al. 2017; Beltrán et al. 2018; Yan, Mao, and Li 2018; Yang, Luo, and Urtasun 2018; Zeng et al. 2018; Zhou and Tuzel 2018; Chen et al. 2019; Lang et al. 2019; Qi et al. 2019; Shi, Wang, and Li 2019; Yang et al. 2019; Zhou et al.…”
3D fully convolutional network for vehicle detection in point cloud (2016, methodology): “ bounding box detection is projected back to the image plane and the minimum rectangle hull of the projection is taken as the 2D bounding boxes. Some previous point cloud based detection methods [2], [16], [30] also use this metric for evaluation. A detection is accepted if the overlap area IoU with”
Multi-view 3D Object Detection Network for Autonomous Driving (2016, methodology): “Existing work usually encodes 3D LIDAR point cloud into a 3D grid [22, 5] or a front view map [14].”
Vote3Deep: Fast object detection in 3D point clouds using efficient convolutional neural networks (2016, methodology): “A CNN-based approach in [7] obtains comparable performance to [5] on KITTI for car detection by projecting the point cloud into a 2D depth map, with an additional channel for the height of a point from the ground.”
Towards Safe Autonomous Driving: Capture Uncertainty in the Deep Neural Network For Lidar 3D Vehicle Detection (2018, methodology): “[5] project range scans into a 2D front-view depth map and use a 2D fully convolutional network (FCN) to detect vehicles.”
A Survey of Autonomous Driving: Common Practices and Emerging Technologies (2019, methodology): “The VeloFCN network [167] proposed to use single-channel depth image as input to a shallow, single-stage convolutional neural”
PIXOR: Real-time 3D Object Detection from Point Clouds (2018, methodology): “As a result, typical systems [6, 37, 20] only run at 1-2 FPS.”
Complex-YOLO: An Euler-Region-Proposal for Real-Time 3D Object Detection on Point Clouds (2018, methodology): “Translation of Point-Clouds into voxels or image stacks by using Convolutional Neural Networks (CNN) [2] [3] [4] [6] [8] [9] [25] [26] 3.”
3D Object Proposals Using Stereo Imagery for Accurate Object Class Detection (2016, neutral): “60 VeloFCNy [54] 71.”
PointPainting: Sequential Fusion for 3D Object Detection (2019, methodology): “The concatenated (painted) lidar points can then be used in any lidar detection method, whether bird’s-eye view [11, 29, 21, 34, 30, 32] or front-view [12, 17].”
IntentNet: Learning to Predict Intention from Raw Sensor Data (2018, methodology): “Approaches that use the projection representation (VeloFCN [15], MV3D [16]) or handle point clouds directly (PointNet [17]) have also been proposed.”
RandLA-Net: Efficient Semantic Segmentation of Large-Scale Point Clouds (2019, background): “To leverage the success of 2D CNNs, many works [30, 8, 63, 27] project/flatten 3D point clouds onto 2D images to address the task of object detection.”
Fusing Bird View LIDAR Point Cloud and Front View Camera Image for Deep Object Detection (2017, methodology): “The DNN-based object detection with either LIDAR [2, 3] or camera [4, 5, 6] data as the input has been widely explored by researchers and pushed to a very high single frame accuracy.”
Stereo R-CNN Based 3D Object Detection for Autonomous Driving (2019, methodology): “[5, 16, 28, 18, 13] project the point cloud into 2D bird’s eye view or front view representations and feed them into the structured convolution network, where [5, 18, 13] exploit fusing multiple LiDAR representations with the RGB image to obtain more dense information.”
Deep Learning for 3D Point Clouds: A Survey (2019, methodology): “pared to [154], 3D FCN-based method [155] obtains a gain of over 20 percent in accuracy, but inevitably costsmore computing resources due to 3D convolutions and the sparsity of the data.”
RTM3D: Real-time Monocular 3D Detection from Object Keypoints for Autonomous Driving (2020, background): “Therefor, most of the recent 3D object detection employ it in different representation to obtain the state-of-the-art model [48, 3, 9, 36, 21].”
Deep-learning in Mobile Robotics - from Perception to Control Systems: A Survey on Why and Why not (2016, methodology): “Li et al. (2016) projected the 3D point cloud from Velodyne to a 2D point map and used a 2D convolutional network to achieve vehicle detection.”
Joint 3D Proposal Generation and Object Detection from View Aggregation (2017, methodology): “When considering sparse and low resolution input such as the Front View [3] or Bird’s Eye View (BEV) [4] point cloud projections, this method is not guaranteed to have enough”