arXiv subjects

Ming Liu

Publications and source records attributed to Ming Liu.

At least 469 records · Page 26Linked to original sources

arXiv2019

Towards Time-Aware Distant Supervision for Relation Extraction

Distant supervision for relation extraction heavily suffers from the wrong labeling problem. To alleviate this issue in news data with the timestamp, we take a new factor time into consideration and propose a novel time-aware distant supervision framework (Time-DS). Time-DS is composed of a time series instance-popularity and two strategies. Instance-popularity is to encode the strong relevance of time and true relation mention. Therefore, instance-popularity would be an effective clue to reduce the noises generated through distant supervision labeling. The two strategies, i.e., hard filter and curriculum learning are both ways to implement instance-popularity for better relation extraction in the manner of Time-DS. The curriculum learning is a more sophisticated and flexible way to exploit instance-popularity to eliminate the bad effects of noises, thus get better relation extraction performance. Experiments on our collected multi-source news corpus show that Time-DS achieves significant improvements for relation extraction.

cs.CL↗

arXiv2019

Using DP Towards A Shortest Path Problem-Related Application

The detection of curved lanes is still challenging for autonomous driving systems. Although current cutting-edge approaches have performed well in real applications, most of them are based on strict model assumptions. Similar to other visual recognition tasks, lane detection can be formulated as a two-dimensional graph searching problem, which can be solved by finding several optimal paths along with line segments and boundaries. In this paper, we present a directed graph model, in which dynamic programming is used to deal with a specific shortest path problem. This model is particularly suitable to represent objects with long continuous shape structure, e.g., lanes and roads. We apply the designed model and proposed an algorithm for detecting lanes by formulating it as the shortest path problem. To evaluate the performance of our proposed algorithm, we tested five sequences (including 1573 frames) from the KITTI database. The results showed that our method achieves an average successful detection precision of 97.5%.

cs.CV↗

arXiv2019

Molybdenum oxide hole selective transport layer by hot wire oxidation-sublimation deposition for silicon heterojunction solar cells

In this article, a novel hot wire oxidation-sublimation deposition (HWOSD) technique was developed to prepare molybdenum oxide (MoOx) thin films with high quality. Silicon heterojunction (SHJ) solar cells with the HWOSD MoOx as a hole selective transport layer (HSL) were fabricated. Thickness of the MoOx layer and annealing process of the solar cells were studied and optimized. A power conversion efficiency up to 21.10% was achieved on a SHJ solar cell using a 14nm MoOx layer as the HSL. Dark current density-voltage-temperature (J-V-T) characteristics of the SHJ solar cell were measured at the temperatures from 200K to 380K. Transport processes including thermionic emission of electrons over the potential barrier and quantum assisted tunneling of holes through the gap states in the MoOx layer were proposed for the MoOx/n c-Si heterojunction. The investigation of the transport mechanisms provides us a better understanding of the characteristics of the novel SHJ solar cells and it is helpful for us to fully demonstrate the potential of such kind of solar cells in the future.

physics.app-ph↗

arXiv2019

PointIT: A Fast Tracking Framework Based on 3D Instance Segmentation

Recently most popular tracking frameworks focus on 2D image sequences. They seldom track the 3D object in point clouds. In this paper, we propose PointIT, a fast, simple tracking method based on 3D on-road instance segmentation. Firstly, we transform 3D LiDAR data into the spherical image with the size of 64 x 512 x 4 and feed it into instance segment model to get the predicted instance mask for each class. Then we use MobileNet as our primary encoder instead of the original ResNet to reduce the computational complexity. Finally, we extend the Sort algorithm with this instance framework to realize tracking in the 3D LiDAR point cloud data. The model is trained on the spherical images dataset with the corresponding instance label masks which are provided by KITTI 3D Object Track dataset. According to the experiment results, our network can achieve on Average Precision (AP) of 0.617 and the performance of multi-tracking task has also been improved.

cs.CV↗

arXiv2019

Neural Speech Synthesis with Transformer Network

Although end-to-end neural text-to-speech (TTS) methods (such as Tacotron2) are proposed and achieve state-of-the-art performance, they still suffer from two problems: 1) low efficiency during training and inference; 2) hard to model long dependency using current recurrent neural networks (RNNs). Inspired by the success of Transformer network in neural machine translation (NMT), in this paper, we introduce and adapt the multi-head attention mechanism to replace the RNN structures and also the original attention mechanism in Tacotron2. With the help of multi-head self-attention, the hidden states in the encoder and decoder are constructed in parallel, which improves the training efficiency. Meanwhile, any two inputs at different times are connected directly by self-attention mechanism, which solves the long range dependency problem effectively. Using phoneme sequences as input, our Transformer TTS network generates mel spectrograms, followed by a WaveNet vocoder to output the final audio results. Experiments are conducted to test the efficiency and performance of our new network. For the efficiency, our Transformer TTS network can speed up the training about 4.25 times faster compared with Tacotron2. For the performance, rigorous human tests show that our proposed model achieves state-of-the-art performance (outperforms Tacotron2 with a gap of 0.048) and is very close to human quality (4.39 vs 4.44 in MOS).

cs.CL↗

arXiv2019

VR-Goggles for Robots: Real-to-sim Domain Adaptation for Visual Control

In this paper, we deal with the reality gap from a novel perspective, targeting transferring Deep Reinforcement Learning (DRL) policies learned in simulated environments to the real-world domain for visual control tasks. Instead of adopting the common solutions to the problem by increasing the visual fidelity of synthetic images output from simulators during the training phase, we seek to tackle the problem by translating the real-world image streams back to the synthetic domain during the deployment phase, to make the robot feel at home. We propose this as a lightweight, flexible, and efficient solution for visual control, as 1) no extra transfer steps are required during the expensive training of DRL agents in simulation; 2) the trained DRL agents will not be constrained to being deployable in only one specific real-world environment; 3) the policy training and the transfer operations are decoupled, and can be conducted in parallel. Besides this, we propose a simple yet effective shift loss that is agnostic to the downstream task, to constrain the consistency between subsequent frames which is important for consistent policy outputs. We validate the shift loss for artistic style transfer for videos and domain adaptation, and validate our visual control approach in indoor and outdoor robotics experiments.

cs.RO↗

arXiv2019

Focal Loss in 3D Object Detection

3D object detection is still an open problem in autonomous driving scenes. When recognizing and localizing key objects from sparse 3D inputs, autonomous vehicles suffer from a larger continuous searching space and higher fore-background imbalance compared to image-based object detection. In this paper, we aim to solve this fore-background imbalance in 3D object detection. Inspired by the recent use of focal loss in image-based object detection, we extend this hard-mining improvement of binary cross entropy to point-cloud-based object detection and conduct experiments to show its performance based on two different 3D detectors: 3D-FCN and VoxelNet. The evaluation results show up to 11.2AP gains through the focal loss in a wide range of hyperparameters for 3D object detection.

cs.CV↗

arXiv2018

Learning Symmetry Consistent Deep CNNs for Face Completion

Deep convolutional networks (CNNs) have achieved great success in face completion to generate plausible facial structures. These methods, however, are limited in maintaining global consistency among face components and recovering fine facial details. On the other hand, reflectional symmetry is a prominent property of face image and benefits face recognition and consistency modeling, yet remaining uninvestigated in deep face completion. In this work, we leverage two kinds of symmetry-enforcing subnets to form a symmetry-consistent CNN model (i.e., SymmFCNet) for effective face completion. For missing pixels on only one of the half-faces, an illumination-reweighted warping subnet is developed to guide the warping and illumination reweighting of the other half-face. As for missing pixels on both of half-faces, we present a generative reconstruction subnet together with a perceptual symmetry loss to enforce symmetry consistency of recovered structures. The SymmFCNet is constructed by stacking generative reconstruction subnet upon illumination-reweighted warping subnet, and can be end-to-end learned from training set of unaligned face images. Experiments show that SymmFCNet can generate high quality results on images with synthetic and real occlusion, and performs favorably against state-of-the-arts.

cs.CV↗

arXiv2018

Semi-Semantic Line-Cluster Assisted Monocular SLAM for Indoor Environments

This paper presents a novel method to reduce the scale drift for indoor monocular simultaneous localization and mapping (SLAM). We leverage the prior knowledge that in the indoor environment, the line segments form tight clusters, e.g. many door frames in a straight corridor are of the same shape, size and orientation, so the same edges of these door frames form a tight line segment cluster. We implement our method in the popular ORB-SLAM2, which also serves as our baseline. In the front end we detect the line segments in each frame and incrementally cluster them in the 3D space. In the back end, we optimize the map imposing the constraint that the line segments of the same cluster should be the same. Experimental results show that our proposed method successfully reduces the scale drift for indoor monocular SLAM.

cs.CV↗

arXiv2018

One-dimensional van der Waals heterostructures

Property by design is one appealing idea in material synthesis but hard to achieve in practice. A recent successful example is the demonstration of van der Waals (vdW) heterostructures,1-3 in which atomic layers are stacked on each other and different ingredients can be combined beyond symmetry and lattice matching. This concept, usually described as a nanoscale Lego blocks, allows to build sophisticated structures layer by layer. However, this concept has been so far limited in two dimensional (2D) materials. Here we show a class of new material where different layers are coaxially (instead of planarly) stacked. As the structure is in one dimensional (1D) form, we name it "1D vdW heterostructures". We demonstrate a 5 nm diameter nanotube consisting of three different materials: an inner conductive carbon nanotube (CNT), a middle insulating hexagonal boron nitride nanotube (BNNT) and an outside semiconducting MoS2 nanotube. As the technique is highly applicable to other materials in the current 2D libraries,4-6 we anticipate our strategy to be a starting point for discovering a class of new semiconducting nanotube materials. A plethora of function-designable 1D heterostructures will appear after the combination of CNTs, BNNTs and semiconducting nanotubes.

cond-mat.mtrl-sci↗

arXiv2018

Point-cloud-based place recognition using CNN feature extraction

This paper proposes a novel point-cloud-based place recognition system that adopts a deep learning approach for feature extraction. By using a convolutional neural network pre-trained on color images to extract features from a range image without fine-tuning on extra range images, significant improvement has been observed when compared to using hand-crafted features. The resulting system is illumination invariant, rotation invariant and robust against moving objects that are unrelated to the place identity. Apart from the system itself, we also bring to the community a new place recognition dataset containing both point cloud and grayscale images covering a full $360^\circ$ environmental view. In addition, the dataset is organized in such a way that it facilitates experimental validation with respect to rotation invariance or robustness against unrelated moving objects separately.

cs.CV↗

arXiv2018

Mobile Robot Localisation and Navigation Using LEGO NXT and Ultrasonic Sensor

Mobile robots are becoming increasingly important both for individuals and industries. Mobile robotic technology is not only utilised by experts in this field but is also very popular among amateurs. However, implementing a mobile robot to perform tasks autonomously can be expensive because of the need for various types of sensors and the high price of robot platforms. Hence, in this paper we present a mobile robot localisation and navigation system which uses a LEGO ultrasonic sensor in an indoor map based on the LEGO MINDSTORM NXT. This provides an affordable and ready-to-use option for most robot fans. In this paper, an effective method is proposed to extract useful information from the distorted readings collected by the ultrasonic sensor. Then, the particle filter is used to localise the robot. After robot's position is estimated, a sampling-based path planning method is proposed for the robot navigation. This method reduces the robot accumulative motion error by minimising robot turning times and covering distances. The robot localisation and navigation algorithms are implemented in MATLAB. Simulation results show an average accuracy between 1 and 3 cm for three different indoor map locations. Furthermore, experiments performed in a real setup show the effectiveness of the proposed methods.

cs.RO↗

arXiv2018

PointSeg: Real-Time Semantic Segmentation Based on 3D LiDAR Point Cloud

In this paper, we propose PointSeg, a real-time end-to-end semantic segmentation method for road-objects based on spherical images. We take 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. To make PointSeg applicable on a mobile system, we build the model based on the light-weight network, SqueezeNet, with several improvements. It maintains a good balance between memory cost and prediction performance. Our model is trained on spherical images and label masks projected from the KITTI 3D object detection dataset. Experiments show that PointSeg can achieve competitive accuracy with 90fps on a single GPU 1080ti. which makes it quite compatible for autonomous driving applications.

cs.CV↗

arXiv2018

PCR-Pro: 3D Sparse and Different Scale Point Clouds Registration and Robust Estimation of Information Matrix For Pose Graph SLAM

For both indoor and outdoor environments, we propose an efficient and novel method for different scales and sparse 3D point clouds registration that cannot be handled by the current popular ICP approaches. Our algorithm efficiently detects the scale difference between point clouds and uses the keyframes to estimate the relative pose for calculating the scale difference. The algorithm applies a filter and computes the final transformation which coverages to a global minimum. The good estimation of transform and scale helps in the calculation of the covariance matrix using a closed form solution efficiently. This covariance between point clouds helps in the estimation of information matrix for pose-graph SLAM.

cs.RO↗

arXiv2018

Multiple Lane Detection Algorithm Based on Optimised Dense Disparity Map Estimation

Lane detection is very important for self-driving vehicles. In recent years, computer stereo vision has been prevalently used to enhance the accuracy of the lane detection systems. This paper mainly presents a multiple lane detection algorithm developed based on optimised dense disparity map estimation, where the disparity information obtained at time t_{n} is utilised to optimise the process of disparity estimation at time t_{n+1}. This is achieved by estimating the road model at time t_{n} and then controlling the search range for the disparity estimation at time t_{n+1}. The lanes are then detected using our previously published algorithm, where the vanishing point information is used to model the lanes. The experimental results illustrate that the runtime of the disparity estimation is reduced by around 37% and the accuracy of the lane detection is about 99%.

cs.CV↗

arXiv2018

Real-Time Subpixel Fast Bilateral Stereo

Stereo vision technique has been widely used in robotic systems to acquire 3-D information. In recent years, many researchers have applied bilateral filtering in stereo vision to adaptively aggregate the matching costs. This has greatly improved the accuracy of the estimated disparity maps. However, the process of filtering the whole cost volume is very time consuming and therefore the researchers have to resort to some powerful hardware for the real-time purpose. This paper presents the implementation of fast bilateral stereo on a state-of-the-art GPU. By highly exploiting the parallel computing architecture of the GPU, the fast bilateral stereo performs in real time when processing the Middlebury stereo datasets.

cs.CV↗

arXiv2018

Fast Rigid 3D Registration Solution: A Simple Method Free of SVD and Eigen-Decomposition

A novel solution is obtained to solve the rigid 3D registration problem, motivated by previous eigen-decomposition approaches. Different from existing solvers, the proposed algorithm does not require sophisticated matrix operations e.g. singular value decomposition or eigenvalue decomposition. Instead, the optimal eigenvector of the point cross-covariance matrix can be computed within several iterations. It is also proven that the optimal rotation matrix can be directly computed for cases without need of quaternion. The simple framework provides very easy approach of integer-implementation on embedded platforms. Simulations on noise-corrupted point clouds have verified the robustness and computation speed of the proposed method. The final results indicate that the proposed algorithm is accurate, robust and owns over $60\% \sim 80\%$ less computation time than representatives. It has also been applied to real-world applications for faster relative robotic navigation.

eess.SY↗

arXiv2018

Curiosity-driven Exploration for Mapless Navigation with Deep Reinforcement Learning

This paper investigates exploration strategies of Deep Reinforcement Learning (DRL) methods to learn navigation policies for mobile robots. In particular, we augment the normal external reward for training DRL algorithms with intrinsic reward signals measured by curiosity. We test our approach in a mapless navigation setting, where the autonomous agent is required to navigate without the occupancy map of the environment, to targets whose relative locations can be easily acquired through low-cost solutions (e.g., visible light localization, Wi-Fi signal localization). We validate that the intrinsic motivation is crucial for improving DRL performance in tasks with challenging exploration requirements. Our experimental results show that our proposed method is able to more effectively learn navigation policies, and has better generalization capabilities in previously unseen environments. A video of our experimental results can be found at https://goo.gl/pWbpcF.

cs.RO↗