arXiv subjects

Xinming Huang

Publications and source records attributed to Xinming Huang.

At least 37 records · Page 2Linked to original sources

arXiv2021

A Divide-and-Merge Point Cloud Clustering Algorithm for LiDAR Panoptic Segmentation

Clustering objects from the LiDAR point cloud is an important research problem with many applications such as autonomous driving. To meet the real-time requirement, existing research proposed to apply the connected-component-labeling (CCL) technique on LiDAR spherical range image with a heuristic condition to check if two neighbor points are connected. However, LiDAR range image is different from a binary image which has a deterministic condition to tell if two pixels belong to the same component. The heuristic condition used on the LiDAR range image only works empirically, which suggests the LiDAR clustering algorithm should be robust to potential failures of the empirical heuristic condition. To overcome this challenge, this paper proposes a divide-and-merge LiDAR clustering algorithm. This algorithm firstly conducts clustering in each evenly divided local region, then merges the local clustered small components by voting on edge point pairs. Assuming there are $N$ LiDAR points of objects in total with $m$ divided local regions, the time complexity of the proposed algorithm is $O(N)+O(m^2)$. A smaller $m$ means the voting will involve more neighbor points, but the time complexity will become larger. So the $m$ controls the trade-off between the time complexity and the clustering accuracy. A proper $m$ helps the proposed algorithm work in real-time as well as maintain good performance. We evaluate the divide-and-merge clustering algorithm on the SemanticKITTI panoptic segmentation benchmark by cascading it with a state-of-the-art semantic segmentation model. The final performance evaluated through the leaderboard achieves the best among all published methods. The proposed algorithm is implemented with C++ and wrapped as a python function. It can be easily used with the modern deep learning framework in python.

cs.CV↗

arXiv2021

FIDNet: LiDAR Point Cloud Semantic Segmentation with Fully Interpolation Decoding

Projecting the point cloud on the 2D spherical range image transforms the LiDAR semantic segmentation to a 2D segmentation task on the range image. However, the LiDAR range image is still naturally different from the regular 2D RGB image; for example, each position on the range image encodes the unique geometry information. In this paper, we propose a new projection-based LiDAR semantic segmentation pipeline that consists of a novel network structure and an efficient post-processing step. In our network structure, we design a FID (fully interpolation decoding) module that directly upsamples the multi-resolution feature maps using bilinear interpolation. Inspired by the 3D distance interpolation used in PointNet++, we argue this FID module is a 2D version distance interpolation on $(θ, ϕ)$ space. As a parameter-free decoding module, the FID largely reduces the model complexity by maintaining good performance. Besides the network structure, we empirically find that our model predictions have clear boundaries between different semantic classes. This makes us rethink whether the widely used K-nearest-neighbor post-processing is still necessary for our pipeline. Then, we realize the many-to-one mapping causes the blurring effect that some points are mapped into the same pixel and share the same label. Therefore, we propose to process those occluded points by assigning the nearest predicted label to them. This NLA (nearest label assignment) post-processing step shows a better performance than KNN with faster inference speed in the ablation study. On the SemanticKITTI dataset, our pipeline achieves the best performance among all projection-based methods with $64 \times 2048$ resolution and all point-wise solutions. With a ResNet-34 as the backbone, both the training and testing of our model can be finished on a single RTX 2080 Ti with 11G memory. The code is released.

cs.CV↗

arXiv2021

A Technical Survey and Evaluation of Traditional Point Cloud Clustering Methods for LiDAR Panoptic Segmentation

LiDAR panoptic segmentation is a newly proposed technical task for autonomous driving. In contrast to popular end-to-end deep learning solutions, we propose a hybrid method with an existing semantic segmentation network to extract semantic information and a traditional LiDAR point cloud cluster algorithm to split each instance object. We argue geometry-based traditional clustering algorithms are worth being considered by showing a state-of-the-art performance among all published end-to-end deep learning solutions on the panoptic segmentation leaderboard of the SemanticKITTI dataset. To our best knowledge, we are the first to attempt the point cloud panoptic segmentation with clustering algorithms. Therefore, instead of working on new models, we give a comprehensive technical survey in this paper by implementing four typical cluster methods and report their performances on the benchmark. Those four cluster methods are the most representative ones with real-time running speed. They are implemented with C++ in this paper and then wrapped as a python function for seamless integration with the existing deep learning frameworks. We release our code for peer researchers who might be interested in this problem.

cs.CV↗

arXiv2021

Revisiting 2D Convolutional Neural Networks for Graph-based Applications

Graph convolutional networks (GCNs) are widely used in graph-based applications such as graph classification and segmentation. However, current GCNs have limitations on implementation such as network architectures due to their irregular inputs. In contrast, convolutional neural networks (CNNs) are capable of extracting rich features from large-scale input data, but they do not support general graph inputs. To bridge the gap between GCNs and CNNs, in this paper we study the problem of how to effectively and efficiently map general graphs to 2D grids that CNNs can be directly applied to, while preserving graph topology as much as possible. We therefore propose two novel graph-to-grid mapping schemes, namely, {\em graph-preserving grid layout (GPGL)} and its extension {\em Hierarchical GPGL (H-GPGL)} for computational efficiency. We formulate the GPGL problem as integer programming and further propose an approximate yet efficient solver based on a penalized Kamada-Kawai method, a well-known optimization algorithm in 2D graph drawing. We propose a novel vertex separation penalty that encourages graph vertices to lay on the grid without any overlap. Along with this image representation, even extra 2D maxpooling layers contribute to the PointNet, a widely applied point-based neural network. We demonstrate the empirical success of GPGL on general graph classification with small graphs and H-GPGL on 3D point cloud segmentation with large graphs, based on 2D CNNs including VGG16, ResNet50 and multi-scale maxout (MSM) CNN.

cs.CV↗

arXiv2021

RoadNet-RT: High Throughput CNN Architecture and SoC Design for Real-Time Road Segmentation

In recent years, convolutional neural network has gained popularity in many engineering applications especially for computer vision. In order to achieve better performance, often more complex structures and advanced operations are incorporated into the neural networks, which results very long inference time. For time-critical tasks such as autonomous driving and virtual reality, real-time processing is fundamental. In order to reach real-time process speed, a light-weight, high-throughput CNN architecture namely RoadNet-RT is proposed for road segmentation in this paper. It achieves 90.33% MaxF score on test set of KITTI road segmentation task and 8 ms per frame when running on GTX 1080 GPU. Comparing to the state-of-the-art network, RoadNet-RT speeds up the inference time by a factor of 20 at the cost of only 6.2% accuracy loss. For hardware design optimization, several techniques such as depthwise separable convolution and non-uniformed kernel size convolution are customized designed to further reduce the processing time. The proposed CNN architecture has been successfully implemented on an FPGA ZCU102 MPSoC platform that achieves the computation capability of 83.05 GOPS. The system throughput reaches 327.9 frames per second with image size 1216x176.

eess.IV↗

arXiv2021

Deep Lucas-Kanade Homography for Multimodal Image Alignment

Estimating homography to align image pairs captured by different sensors or image pairs with large appearance changes is an important and general challenge for many computer vision applications. In contrast to others, we propose a generic solution to pixel-wise align multimodal image pairs by extending the traditional Lucas-Kanade algorithm with networks. The key contribution in our method is how we construct feature maps, named as deep Lucas-Kanade feature map (DLKFM). The learned DLKFM can spontaneously recognize invariant features under various appearance-changing conditions. It also has two nice properties for the Lucas-Kanade algorithm: (1) The template feature map keeps brightness consistency with the input feature map, thus the color difference is very small while they are well-aligned. (2) The Lucas-Kanade objective function built on DLKFM has a smooth landscape around ground truth homography parameters, so the iterative solution of the Lucas-Kanade can easily converge to the ground truth. With those properties, directly updating the Lucas-Kanade algorithm on our feature maps will precisely align image pairs with large appearance changes. We share the datasets, code, and demo video online.

cs.CV↗

arXiv2021

A Surface Geometry Model for LiDAR Depth Completion

LiDAR depth completion is a task that predicts depth values for every pixel on the corresponding camera frame, although only sparse LiDAR points are available. Most of the existing state-of-the-art solutions are based on deep neural networks, which need a large amount of data and heavy computations for training the models. In this letter, a novel non-learning depth completion method is proposed by exploiting the local surface geometry that is enhanced by an outlier removal algorithm. The proposed surface geometry model is inspired by the observation that most pixels with unknown depth have a nearby LiDAR point. Therefore, it is assumed those pixels share the same surface with the nearest LiDAR point, and their respective depth can be estimated as the nearest LiDAR depth value plus a residual error. The residual error is calculated by using a derived equation with several physical parameters as input, including the known camera intrinsic parameters, estimated normal vector, and offset distance on the image plane. The proposed method is further enhanced by an outlier removal algorithm that is designed to remove incorrectly mapped LiDAR points from occluded regions. On KITTI dataset, the proposed solution achieves the best error performance among all existing non-learning methods and is comparable to the best self-supervised learning method and some supervised learning methods. Moreover, since outlier points from occluded regions is a commonly existing problem, the proposed outlier removal algorithm is a general preprocessing step that is applicable to many robotic systems with both camera and LiDAR sensors.

cs.CV↗

arXiv2020

A Survey on 3D LiDAR Localization for Autonomous Vehicles

LiDAR sensors are becoming one of the most essential sensors in achieving full autonomy for self driving cars. LiDARs are able to produce rich, dense and precise spatial data, which can tremendously help in localizing and tracking a moving vehicle. In this paper, we review the latest finding in 3D LiDAR localization for autonomous driving cars, and analyse the results obtained by each method, in an effort to guide the research community towards the path that seems to be the most promising.

cs.CV↗

arXiv2020

Learning to Segment 3D Point Clouds in 2D Image Space

In contrast to the literature where local patterns in 3D point clouds are captured by customized convolutional operators, in this paper we study the problem of how to effectively and efficiently project such point clouds into a 2D image space so that traditional 2D convolutional neural networks (CNNs) such as U-Net can be applied for segmentation. To this end, we are motivated by graph drawing and reformulate it as an integer programming problem to learn the topology-preserving graph-to-grid mapping for each individual point cloud. To accelerate the computation in practice, we further propose a novel hierarchical approximate algorithm. With the help of the Delaunay triangulation for graph construction from point clouds and a multi-scale U-Net for segmentation, we manage to demonstrate the state-of-the-art performance on ShapeNet and PartNet, respectively, with significant improvement over the literature. Code is available at https://github.com/Zhang-VISLab.

cs.CV↗

arXiv2020

TreeRNN: Topology-Preserving Deep GraphEmbedding and Learning

General graphs are difficult for learning due to their irregular structures. Existing works employ message passing along graph edges to extract local patterns using customized graph kernels, but few of them are effective for the integration of such local patterns into global features. In contrast, in this paper we study the methods to transfer the graphs into trees so that explicit orders are learned to direct the feature integration from local to global. To this end, we apply the breadth first search (BFS) to construct trees from the graphs, which adds direction to the graph edges from the center node to the peripheral nodes. In addition, we proposed a novel projection scheme that transfer the trees to image representations, which is suitable for conventional convolution neural networks (CNNs) and recurrent neural networks (RNNs). To best learn the patterns from the graph-tree-images, we propose TreeRNN, a 2D RNN architecture that recurrently integrates the image pixels by rows and columns to help classify the graph categories. We evaluate the proposed method on several graph classification datasets, and manage to demonstrate comparable accuracy with the state-of-the-art on MUTAG, PTC-MR and NCI1 datasets.

cs.CV↗

arXiv2020

DepthNet: Real-Time LiDAR Point Cloud Depth Completion for Autonomous Vehicles

Autonomous vehicles rely heavily on sensors such as camera and LiDAR, which provide real-time information about their surroundings for the tasks of perception, planning and control. Typically a LiDAR can only provide sparse point cloud owing to a limited number of scanning lines. By employing depth completion, a dense depth map can be generated by assigning each camera pixel a corresponding depth value. However, the existing depth completion convolutional neural networks are very complex that requires high-end GPUs for processing, and thus they are not applicable to real-time autonomous driving. In this paper, a light-weight network is proposed for the task of LiDAR point cloud depth completion. With an astonishing 96.2% reduction in the number of parameters, it still achieves comparable performance (9.3% better in MAE but 3.9% worse in RMSE) to the state-of-the-art network. For real-time embedded platforms, depthwise separable technique is applied to both convolution and deconvolution operations and the number of parameters decreases further by a factor of 7.3, with only a small percentage increase in RMSE and MAE performance. Moreover, a system-on-chip architecture for depth completion is developed on a PYNQ-based FPGA platform that achieves real-time processing for HDL-64E LiDAR at the speed 11.1 frame per second.

eess.IV↗

arXiv2020

Automatic Building and Labeling of HD Maps with Deep Learning

In a world where autonomous driving cars are becoming increasingly more common, creating an adequate infrastructure for this new technology is essential. This includes building and labeling high-definition (HD) maps accurately and efficiently. Today, the process of creating HD maps requires a lot of human input, which takes time and is prone to errors. In this paper, we propose a novel method capable of generating labelled HD maps from raw sensor data. We implemented and tested our methods on several urban scenarios using data collected from our test vehicle. The results show that the pro-posed deep learning based method can produce highly accurate HD maps. This approach speeds up the process of building and labeling HD maps, which can make meaningful contribution to the deployment of autonomous vehicle.

cs.CV↗

arXiv2020

Pedestrian Tracking with Gated Recurrent Units and Attention Mechanisms

Pedestrian tracking has long been considered an important problem, especially in security applications. Previously,many approaches have been proposed with various types of sensors. One popular method is Pedestrian Dead Reckoning(PDR) [1] which is based on the inertial measurement unit(IMU) sensor. However PDR is an integration and threshold based method, which suffers from accumulation errors and low accuracy. In this paper, we propose a novel method in which the sensor data is fed into a deep learning model to predict the displacements and orientations of the pedestrian. We also devise a new apparatus to collect and construct databases containing synchronized IMU sensor data and precise locations measured by a LIDAR. The preliminary results are promising, and we plan to push this forward by collecting more data and adapting the deep learning model for all general pedestrian motions.

cs.CV↗

arXiv2020

PointNet on FPGA for Real-Time LiDAR Point Cloud Processing

LiDAR sensors have been widely used in many autonomous vehicle modalities, such as perception, mapping, and localization. This paper presents an FPGA-based deep learning platform for real-time point cloud processing targeted on autonomous vehicles. The software driver for the Velodyne LiDAR sensor is modified and moved into the on-chip processor system, while the programmable logic is designed as a customized hardware accelerator. As the state-of-art deep learning algorithm for point cloud processing, PointNet is successfully implemented on the proposed FPGA platform. Targeted on a Xilinx Zynq UltraScale+ MPSoC ZCU104 development board, the FPGA implementations of PointNet achieve the computing performance of 182.1 GOPS and 280.0 GOPS for classification and segmentation respectively. The proposed design can support an input up to 4096 points per frame. The processing time is 19.8 ms for classification and 34.6 ms for segmentation, which meets the real-time requirement for most of the existing LiDAR sensors.

eess.SP↗

arXiv2020

A Unified Hardware Architecture for Convolutions and Deconvolutions in CNN

In this paper, a scalable neural network hardware architecture for image segmentation is proposed. By sharing the same computing resources, both convolution and deconvolution operations are handled by the same process element array. In addition, access to on-chip and off-chip memories is optimized to alleviate the burden introduced by partial sum. As an example, SegNet-Basic has been implemented using the proposed unified architecture by targeting on Xilinx ZC706 FPGA, which achieves the performance of 151.5 GOPS and 94.3 GOPS for convolution and deconvolution respectively. This unified convolution/deconvolution design is applicable to other CNNs with deconvolution.

eess.SP↗

arXiv2019

Graph-Preserving Grid Layout: A Simple Graph Drawing Method for Graph Classification using CNNs

Graph convolutional networks (GCNs) suffer from the irregularity of graphs, while more widely-used convolutional neural networks (CNNs) benefit from regular grids. To bridge the gap between GCN and CNN, in contrast to previous works on generalizing the basic operations in CNNs to graph data, in this paper we address the problem of how to project undirected graphs onto the grid in a {\em principled} way where CNNs can be used as backbone for geometric deep learning. To this end, inspired by the literature of graph drawing we propose a novel graph-preserving grid layout (GPGL), an integer programming that minimizes the topological loss on the grid. Technically we propose solving GPGL approximately using a {\em regularized} Kamada-Kawai algorithm, a well-known nonconvex optimization technique in graph drawing, with a vertex separation penalty that improves the rounding performance on top of the solutions from relaxation. Using GPGL we can easily conduct data augmentation as every local minimum will lead to a grid layout for the same graph. Together with the help of multi-scale maxout CNNs, we demonstrate the empirical success of our method for graph classification.

cs.LG↗

arXiv2019

Maximum A Posteriori Probability (MAP) Joint Fine Frequency Offset and Channel Estimation for MIMO Systems with Channels of Arbitrary Correlation

Channel and frequency offset estimation is a classic topic with a large body of prior work using mainly maximum likelihood (ML) approach together with Cramér-Rao Lower bounds (CRLB) analysis. We provide the maximum a posteriori (MAP) estimation solution which is particularly useful for for tracking where previous estimation can be used as prior knowledge. Unlike the ML cases, the corresponding Bayesian Cramér-Rao Lower bound (BCRLB) shows clear relation with parameters and a low complexity algorithm achieves the BCRLB in almost all SNR range. We allow the time invariant channel within a packet to have arbitrary correlation and mean. The estimation is based on pilot/training signals. An unexpected result is that the joint MAP estimation is equivalent to an individual MAP estimation of the frequency offset first, again different from the ML results. We provide insight on the pilot/training signal design based on the BCRLB. Unlike past algorithms that trade performance and/or complexity for the accommodation of time varying channels, the MAP solution provides a different route for dealing with time variation. Within a short enough (segment of) packet where the channel and CFO are approximately time invariant, the low complexity algorithm can be employed. Similar to belief propagation, the estimation of the previous (segment of) packet can serve as the prior knowledge for the next (segment of) packet.

eess.SP↗

arXiv2019

An Interactive LiDAR to Camera Calibration

Recent progress in the automated driving system (ADS) and advanced driver assistant system (ADAS) has shown that the combined use of 3D light detection and ranging (LiDAR) and the camera is essential for an intelligent vehicle to perceive and understand its surroundings. LiDAR-camera fusion requires precise intrinsic and extrinsic calibrations between the sensors. However, due to the limitation of the calibration equipment and susceptibility to noise, algorithms in existing methods tend to fail in finding LiDAR-camera correspondences in long-range. In this paper, we introduced an interactive LiDAR to camera calibration toolbox to estimate the intrinsic and extrinsic transforms. This toolbox automatically detects the corner of a planer board from a sequence of LiDAR frames and provides a convenient user interface for annotating the corresponding pixels on camera frames. Since the toolbox only detects the top corner of the board, there is no need to prepare a precise polygon planar board or a checkerboard with different reflectivity areas as in the existing methods. Furthermore, the toolbox uses genetic algorithms to estimate the transforms and supports multiple camera models such as the pinhole camera model and the fisheye camera model. Experiments using Velodyne VLP-16 LiDAR and Point Grey Chameleon 3 camera show robust results.

eess.IV↗