arXiv subjects

Bin Yang

Publications and source records attributed to Bin Yang.

At least 361 records · Page 20Linked to original sources

arXiv2020

Force myography benchmark data for hand gesture recognition and transfer learning

Force myography has recently gained increasing attention for hand gesture recognition tasks. However, there is a lack of publicly available benchmark data, with most existing studies collecting their own data often with custom hardware and for varying sets of gestures. This limits the ability to compare various algorithms, as well as the possibility for research to be done without first needing to collect data oneself. We contribute to the advancement of this field by making accessible a benchmark dataset collected using a commercially available sensor setup from 20 persons covering 18 unique gestures, in the hope of allowing further comparison of results as well as easier entry into this field of research. We illustrate one use-case for such data, showing how we can improve gesture recognition accuracy by utilising transfer learning to incorporate data from multiple other persons. This also illustrates that the dataset can serve as a benchmark dataset to facilitate research on transfer learning algorithms.

cs.LG↗

arXiv2020

RadarNet: Exploiting Radar for Robust Perception of Dynamic Objects

We tackle the problem of exploiting Radar for perception in the context of self-driving as Radar provides complementary information to other sensors such as LiDAR or cameras in the form of Doppler velocity. The main challenges of using Radar are the noise and measurement ambiguities which have been a struggle for existing simple input or output fusion methods. To better address this, we propose a new solution that exploits both LiDAR and Radar sensors for perception. Our approach, dubbed RadarNet, features a voxel-based early fusion and an attention-based late fusion, which learn from data to exploit both geometric and dynamic information of Radar data. RadarNet achieves state-of-the-art results on two large-scale real-world datasets in the tasks of object detection and velocity estimation. We further show that exploiting Radar improves the perception capabilities of detecting faraway objects and understanding the motion of dynamic objects.

cs.CV↗

arXiv2020

Learning Lane Graph Representations for Motion Forecasting

We propose a motion forecasting model that exploits a novel structured map representation as well as actor-map interactions. Instead of encoding vectorized maps as raster images, we construct a lane graph from raw map data to explicitly preserve the map structure. To capture the complex topology and long range dependencies of the lane graph, we propose LaneGCN which extends graph convolutions with multiple adjacency matrices and along-lane dilation. To capture the complex interactions between actors and maps, we exploit a fusion network consisting of four types of interactions, actor-to-lane, lane-to-lane, lane-to-actor and actor-to-actor. Powered by LaneGCN and actor-map interactions, our model is able to predict accurate and realistic multi-modal trajectories. Our approach significantly outperforms the state-of-the-art on the large scale Argoverse motion forecasting benchmark.

cs.CV↗

arXiv2020

PnPNet: End-to-End Perception and Prediction with Tracking in the Loop

We tackle the problem of joint perception and motion forecasting in the context of self-driving vehicles. Towards this goal we propose PnPNet, an end-to-end model that takes as input sequential sensor data, and outputs at each time step object tracks and their future trajectories. The key component is a novel tracking module that generates object tracks online from detections and exploits trajectory level features for motion forecasting. Specifically, the object tracks get updated at each time step by solving both the data association problem and the trajectory estimation problem. Importantly, the whole model is end-to-end trainable and benefits from joint optimization of all tasks. We validate PnPNet on two large-scale driving datasets, and show significant improvements over the state-of-the-art with better occlusion recovery and more accurate future prediction.

cs.CV↗

arXiv2020

LiDARsim: Realistic LiDAR Simulation by Leveraging the Real World

We tackle the problem of producing realistic simulations of LiDAR point clouds, the sensor of preference for most self-driving vehicles. We argue that, by leveraging real data, we can simulate the complex world more realistically compared to employing virtual worlds built from CAD/procedural models. Towards this goal, we first build a large catalog of 3D static maps and 3D dynamic objects by driving around several cities with our self-driving fleet. We can then generate scenarios by selecting a scene from our catalog and "virtually" placing the self-driving vehicle (SDV) and a set of dynamic objects from the catalog in plausible locations in the scene. To produce realistic simulations, we develop a novel simulator that captures both the power of physics-based and learning-based simulation. We first utilize ray casting over the 3D scene and then use a deep neural network to produce deviations from the physics-based simulation, producing realistic LiDAR point clouds. We showcase LiDARsim's usefulness for perception algorithms-testing on long-tail events and end-to-end closed-loop evaluation on safety-critical scenarios.

cs.CV↗

arXiv2020

AeGAN: Time-Frequency Speech Denoising via Generative Adversarial Networks

Automatic speech recognition (ASR) systems are of vital importance nowadays in commonplace tasks such as speech-to-text processing and language translation. This created the need for an ASR system that can operate in realistic crowded environments. Thus, speech enhancement is a valuable building block in ASR systems and other applications such as hearing aids, smartphones and teleconferencing systems. In this paper, a generative adversarial network (GAN) based framework is investigated for the task of speech enhancement, more specifically speech denoising of audio tracks. A new architecture based on CasNet generator and an additional feature-based loss are incorporated to get realistically denoised speech phonetics. Finally, the proposed framework is shown to outperform other learning and traditional model-based speech enhancement approaches.

eess.AS↗

arXiv2020

Physically Realizable Adversarial Examples for LiDAR Object Detection

Modern autonomous driving systems rely heavily on deep learning models to process point cloud sensory data; meanwhile, deep models have been shown to be susceptible to adversarial attacks with visually imperceptible perturbations. Despite the fact that this poses a security concern for the self-driving industry, there has been very little exploration in terms of 3D perception, as most adversarial attacks have only been applied to 2D flat images. In this paper, we address this issue and present a method to generate universal 3D adversarial objects to fool LiDAR detectors. In particular, we demonstrate that placing an adversarial object on the rooftop of any target vehicle to hide the vehicle entirely from LiDAR detectors with a success rate of 80%. We report attack results on a suite of detectors using various input representation of point clouds. We also conduct a pilot study on adversarial defense using data augmentation. This is one step closer towards safer self-driving under unseen conditions from limited training data.

cs.CV↗

arXiv2020

Organ-based Chronological Age Estimation based on 3D MRI Scans

Individuals age differently depending on a multitude of different factors such as lifestyle, medical history and genetics. Often, the global chronological age is not indicative of the true ageing process. An organ-based age estimation would yield a more accurate health state assessment. In this work, we propose a new deep learning architecture for organ-based age estimation based on magnetic resonance images (MRI). The proposed network is a 3D convolutional neural network (CNN) with increased depth and width made possible by the hybrid utilization of inception and fire modules. We apply the proposed framework for the tasks of brain and knee age estimation. Quantitative comparisons against concurrent MR-based regression networks and different 2D and 3D data feeding strategies illustrated the superior performance of the proposed work.

eess.IV↗

arXiv2020

SELD-TCN: Sound Event Localization & Detection via Temporal Convolutional Networks

The understanding of the surrounding environment plays a critical role in autonomous robotic systems, such as self-driving cars. Extensive research has been carried out concerning visual perception. Yet, to obtain a more complete perception of the environment, autonomous systems of the future should also take acoustic information into account. Recent sound event localization and detection (SELD) frameworks utilize convolutional recurrent neural networks (CRNNs). However, considering the recurrent nature of CRNNs, it becomes challenging to implement them efficiently on embedded hardware. Not only are their computations strenuous to parallelize, but they also require high memory bandwidth and large memory buffers. In this work, we develop a more robust and hardware-friendly novel architecture based on a temporal convolutional network(TCN). The proposed framework (SELD-TCN) outperforms the state-of-the-art SELDnet performance on four different datasets. Moreover, SELD-TCN achieves 4x faster training time per epoch and 40x faster inference time on an ordinary graphics processing unit (GPU).

eess.AS↗

arXiv2020

Infinitely Wide Graph Convolutional Networks: Semi-supervised Learning via Gaussian Processes

Graph convolutional neural networks~(GCNs) have recently demonstrated promising results on graph-based semi-supervised classification, but little work has been done to explore their theoretical properties. Recently, several deep neural networks, e.g., fully connected and convolutional neural networks, with infinite hidden units have been proved to be equivalent to Gaussian processes~(GPs). To exploit both the powerful representational capacity of GCNs and the great expressive power of GPs, we investigate similar properties of infinitely wide GCNs. More specifically, we propose a GP regression model via GCNs~(GPGC) for graph-based semi-supervised learning. In the process, we formulate the kernel matrix computation of GPGC in an iterative analytical form. Finally, we derive a conditional distribution for the labels of unobserved nodes based on the graph structure, labels for the observed nodes, and the feature matrix of all the nodes. We conduct extensive experiments to evaluate the semi-supervised classification performance of GPGC and demonstrate that it outperforms other state-of-the-art methods by a clear margin on all the datasets while being efficient.

cs.LG↗

arXiv2020

ipA-MedGAN: Inpainting of Arbitrary Regions in Medical Imaging

Local deformations in medical modalities are common phenomena due to a multitude of factors such as metallic implants or limited field of views in magnetic resonance imaging (MRI). Completion of the missing or distorted regions is of special interest for automatic image analysis frameworks to enhance post-processing tasks such as segmentation or classification. In this work, we propose a new generative framework for medical image inpainting, titled ipA-MedGAN. It bypasses the limitations of previous frameworks by enabling inpainting of arbitrary shaped regions without a prior localization of the regions of interest. Thorough qualitative and quantitative comparisons with other inpainting and translational approaches have illustrated the superior performance of the proposed framework for the task of brain MR inpainting.

eess.IV↗

arXiv2019

Isospin violating decay $D^*_s \to D_s π^0$ in chiral perturbation theory

We systematically calculate the isospin violating decay, $D_s^*\to D_sπ^0$, with the heavy meson chiral perturbation theory up to $\mathcal{O}(p^3)$ including the loop diagrams. The $\mathcal{O}(p^3)$ tree level amplitudes contain four undetermined LECs. We use two strategies to estimate them. With the nonanalytic dominance approximation, we get $Γ[D_s^\ast\to D_sπ^0]=(3.38\pm0.12)$ eV. With the naturalness assumption, we give a possible range of the isospin violating decay width, $[1.11-6.88]$ eV. We find that the contribution of the $\mathcal{O}(p^3)$ corrections might be significant.

hep-ph↗

arXiv2019

Searching for water ice in the coma of interstellar object 2I/Borisov

Interstellar Objects (ISO) passing through our Solar System offer a rare opportunity to probe the physical and chemical processes involved in solid body and planet formation in extrasolar systems. The main objective of our study is to search for diagnostic absorption features of water ice in the near infrared (NIR) spectrum of the second interstellar object 2I/2019 Q4 (Borisov) and compare its ice features to those of the Solar system icy objects. We observed 2I in the NIR on three separate occasions. The first observation was made on 2019 September 19 UT using the SpeX spectrograph at the 3-m IRTF and again on September 24 UT with the GNIRS spectrograph at the 8-m GEMINI telescope and the last observation was made on October 09 UT with IRTF. The spectra obtained from all three nights appear featureless. No absorption features associated with water ice are detected. Spectral modeling suggests that water grains, if present, comprise no more than 10% of the coma cross-section. The comet consistently exhibits a red D-type like spectrum with a spectral slope of about 6% per 100nm, which is similar to that of 1I/'Oumuamua and is comparable to Solar system comets.

astro-ph.EP↗

arXiv2019

Open-Set Recognition Using Intra-Class Splitting

This paper proposes a method to use deep neural networks as end-to-end open-set classifiers. It is based on intra-class data splitting. In open-set recognition, only samples from a limited number of known classes are available for training. During inference, an open-set classifier must reject samples from unknown classes while correctly classifying samples from known classes. The proposed method splits given data into typical and atypical normal subsets by using a closed-set classifier. This enables to model the abnormal classes by atypical normal samples. Accordingly, the open-set recognition problem is reformulated into a traditional classification problem. In addition, a closed-set regularization is proposed to guarantee a high closed-set classification performance. Intensive experiments on five well-known image datasets showed the effectiveness of the proposed method which outperformed the baselines and achieved a distinct improvement over the state-of-the-art methods.

cs.LG↗

arXiv2019

One-Class Feature Learning Using Intra-Class Splitting

This paper proposes a novel generic one-class feature learning method based on intra-class splitting. In one-class classification, feature learning is challenging, because only samples of one class are available during training. Hence, state-of-the-art methods require reference multi-class datasets to pretrain feature extractors. In contrast, the proposed method realizes feature learning by splitting the given normal class into typical and atypical normal samples. By introducing closeness loss and dispersion loss, an intra-class joint training procedure between the two subsets after splitting enables the extraction of valuable features for one-class classification. Various experiments on three well-known image classification datasets demonstrate the effectiveness of our method which outperformed other baseline models in average.

cs.LG↗

arXiv2019

Unsupervised Adversarial Correction of Rigid MR Motion Artifacts

Motion is one of the main sources for artifacts in magnetic resonance (MR) images. It can have significant consequences on the diagnostic quality of the resultant scans. Previously, supervised adversarial approaches have been suggested for the correction of MR motion artifacts. However, these approaches suffer from the limitation of required paired co-registered datasets for training which are often hard or impossible to acquire. Building upon our previous work, we introduce a new adversarial framework with a new generator architecture and loss function for the unsupervised correction of severe rigid motion artifacts in the brain region. Quantitative and qualitative comparisons with other supervised and unsupervised translation approaches showcase the enhanced performance of the introduced framework.

eess.IV↗

arXiv2019

Detection of CN gas in Interstellar Object 2I/Borisov

The detection of Interstellar Objects passing through the Solar System offers the promise of constraining the physical and chemical processes involved in planetary formation in other extrasolar systems. While the effect of outgassing by 1I/2017 U1 ('Oumuamua) was dynamically observed, no direct detection of the ejected material was made. The discovery of the active interstellar comet 2I/Borisov means spectroscopic investigations of the sublimated ices is possible for this object. We report the first detection of gas emitted by an interstellar comet via the near-UV emission of CN from 2I/Borisov at a heliocentric distance of $r$ = 2.7 au on 2019 September 20. The production rate was found to be Q(CN) = $(3.7\pm0.4)\times10^{24}$ s$^{-1}$, using a simple Haser model with an outflow velocity of 0.5 km s$^{-1}$. No other emission was detected, with an upper limit to the production rate of C$_2$ of $4\times10^{24}$ s$^{-1}$. The spectral reflectance slope of the dust coma over $3900$ Å $< λ< 6000$ Å\ is steeper than at longer wavelengths, as found for other comets. Broad band $R_c$ photometry on 2019 September 19 gave a dust production rate of $Afρ=143\pm10$ cm. Modelling of the observed gas and dust production rates constrains the nuclear radius to $0.7-3.3$ km assuming reasonable nuclear properties. Overall, we find the gas, dust and nuclear properties for the first active Interstellar Object are similar to normal Solar System comets.

astro-ph.EP↗

arXiv2019

Obtaining a History of the Flux of Cosmic Rays using In Situ Cosmogenic $^{14}$C Trapped in Polar Ice

Carbon-14 ($^{14}$C) is produced in the atmosphere when neutrons from cosmic-ray air showers are captured by $^{14}$N nuclei. Atmospheric $^{14}$C becomes trapped in air bubbles in polar ice as compacted snow (firn) transforms into ice. $^{14}$C is also produced in situ in ice grains by penetrating cosmic-ray neutrons and muons. Recent ice core measurements indicate that in the $^{14}$CO phase, the $^{14}$C is dominated by the in situ cosmogenic component at most ice coring sites. Thus, it should be possible to use ice-bound $^{14}$CO to reconstruct the historical flux of cosmic rays at Earth, without the transport and deposition uncertainties associated with $^{10}$Be or the carbon cycle uncertainties affecting atmospheric $^{14}$CO$_2$. The measurements will be sensitive to the cosmic-ray flux above the energy range most affected by solar modulation. We present estimates of the expected sensitivity of $^{14}$CO in ice cores to the historical flux of Galactic cosmic rays, based on recent studies of $^{14}$CO in polar ice.

astro-ph.HE↗