arXiv subjects

Yubo Zhang

Publications and source records attributed to Yubo Zhang.

82 records · Page 5Linked to original sources

arXiv2019

A Structured Model For Action Detection

A dominant paradigm for learning-based approaches in computer vision is training generic models, such as ResNet for image recognition, or I3D for video understanding, on large datasets and allowing them to discover the optimal representation for the problem at hand. While this is an obviously attractive approach, it is not applicable in all scenarios. We claim that action detection is one such challenging problem - the models that need to be trained are large, and labeled data is expensive to obtain. To address this limitation, we propose to incorporate domain knowledge into the structure of the model, simplifying optimization. In particular, we augment a standard I3D network with a tracking module to aggregate long term motion patterns, and use a graph convolutional network to reason about interactions between actors and objects. Evaluated on the challenging AVA dataset, the proposed approach improves over the I3D baseline by 5.5% mAP and over the state-of-the-art by 4.8% mAP.

cs.CV↗

arXiv2019

Towards Safety-Aware Computing System Design in Autonomous Vehicles

Recently, autonomous driving development ignited competition among car makers and technical corporations. Low-level automation cars are already commercially available. But high automated vehicles where the vehicle drives by itself without human monitoring is still at infancy. Such autonomous vehicles (AVs) rely on the computing system in the car to to interpret the environment and make driving decisions. Therefore, computing system design is essential particularly in enhancing the attainment of driving safety. However, to our knowledge, no clear guideline exists so far regarding safety-aware AV computing system and architecture design. To understand the safety requirement of AV computing system, we performed a field study by running industrial Level-4 autonomous driving fleets in various locations, road conditions, and traffic patterns. The field study indicates that traditional computing system performance metrics, such as tail latency, average latency, maximum latency, and timeout, cannot fully satisfy the safety requirement for AV computing system design. To address this issue, we propose a `safety score' as a primary metric for measuring the level of safety in AV computing system design. Furthermore, we propose a perception latency model, which helps architects estimate the safety score of given architecture and system design without physically testing them in an AV. We demonstrate the use of our safety score and latency model, by developing and evaluating a safety-aware AV computing system computation hardware resource management scheme.

cs.RO↗

arXiv2019

Identification of a monoclinic metallic state in VO2 from a modified first-principles approach

Metal-insulator transition underlies many remarkable and technologically important phenomena in VO2. Even though its monoclinic structure had before been the reserve of the insulating state, recent experiments have observed an unexpected monoclinic metallic state. Here we use a modified approach combining first-principles calculations with orbital-biased potentials to reproduce the correct stability ordering and electronic structure of different phases of VO2. We identify a ferromagnetic monoclinic metal that is likely to be the experimentally observed mysterious metastable state. Furthermore, our calculations show that an isostructural insulator-metal electronic transition is followed by the lattice distortion from the monoclinic structure to the rutile structure. These results not only explain the experimental observations of the monoclinic metallic state and the decoupled structural and electronic transitions of VO2, but also provide a understanding for the metal-insulator transition in other strongly correlated d electron systems.

cond-mat.str-el↗

arXiv2018

Landscape of competing stripe and magnetic phases in cuprates

Realistic modeling of competing phases in complex quantum materials has proven extremely challenging. For example, much of the existing density-functional-theory-based first-principles framework fails in the cuprate superconductors. Various many-body approaches involve generic model Hamiltonians and do not account for the couplings between spin, charge, and lattice. Here, by deploying the recently constructed strongly-constrained-and-appropriately-normed density functional, we show how landscapes of competing stripe and magnetic phases can be addressed on a first-principles basis in YBa2Cu3O6 and YBa2Cu3O7 as archetype cuprate compounds. We invoke no free parameters such as the Hubbard U, which has been the basis of much of the cuprate literature. Lattice degrees of freedom are found to be crucially important in stabilizing the various phases.

cond-mat.supr-con↗

arXiv2018

Antiferromagnetic Ground State of La$_{2}$CuO$_{4}$: A Parameter-free Ab Initio Description

We show how an accurate first-principles treatment of the antiferromagnetic (AFM) ground state of La$_2$CuO$_4$ can be obtained without invoking any free parameters such as the Hubbard $U$. The magnitude and orientation of our theoretically predicted magnetic moment of $0.495 μ_{B}$ on Cu-sites along the (100) direction are in excellent accord with experimental results. The computed values of the band gap (1.00 eV) and the exchange-coupling (-138 meV) match the corresponding experimental values. We identify interesting band splittings below the Fermi energy, including an appreciable Hund's splitting of 1.25 eV. The magnetic form factor obtained from neutron scattering experiments is also well described by our calculations. Our study opens up a new pathway for first-principles investigations of electronic and atomic structures and phase diagrams of cuprates and other complex materials.

cond-mat.str-el↗

arXiv2018

Gesture-based Bootstrapping for Egocentric Hand Segmentation

Accurately identifying hands in images is a key sub-task for human activity understanding with wearable first-person point-of-view cameras. Traditional hand segmentation approaches rely on a large corpus of manually labeled data to generate robust hand detectors. However, these approaches still face challenges as the appearance of the hand varies greatly across users, tasks, environments or illumination conditions. A key observation in the case of many wearable applications and interfaces is that, it is only necessary to accurately detect the user's hands in a specific situational context. Based on this observation, we introduce an interactive approach to learn a person-specific hand segmentation model that does not require any manually labeled training data. Our approach proceeds in two steps, an interactive bootstrapping step for identifying moving hand regions, followed by learning a personalized user specific hand appearance model. Concretely, our approach uses two convolutional neural networks: (1) a gesture network that uses pre-defined motion information to detect the hand region; and (2) an appearance network that learns a person specific model of the hand region based on the output of the gesture network. During training, to make the appearance network robust to errors in the gesture network, the loss function of the former network incorporates the confidence of the gesture network while learning. Experiments demonstrate the robustness of our approach with an F1 score over 0.8 on all challenging datasets across a wide range of illumination and hand appearance variations, improving over a baseline approach by over 10%.

cs.CV↗

arXiv2017

Comparative first-principles studies of prototypical ferroelectric materials by LDA, GGA, and SCAN meta-GGA

Originating from a broken spatial inversion symmetry, ferroelectricity is a functionality of materials with an electric dipole that can be switched by external electric fields. Spontaneous polarization is a crucial ferroelectric property, and its amplitude is determined by the strength of polar structural distortions. Density functional theory (DFT) is one of the most widely used theoretical methods to study ferroelectric properties, yet it is limited by the levels of approximations in electron exchange-correlation. On the one hand, the local density approximation (LDA) is considered to be more accurate for the conventional perovskite ferroelectrics such as BaTiO3 and PbTiO3 than the generalized gradient approximation (GGA), which suffers from the so-called super-tetragonality error. On the other hand, GGA is more suitable for hydrogen-bonded ferroelectrics than LDA, which largely overestimates the strength of hydrogen bonding in general. We show here that the recently developed general-purpose strongly constrained and appropriately normed (SCAN) meta-GGA functional significantly improves over the traditional LDA/GGA for structural, electric, and energetic properties of diversely-bonded ferroelectric materials with a comparable computational effort, and thus enhances largely the predictive power of DFT in studies of ferroelectric materials. We also address the observed system-dependent performances of LDA and GGA for ferroelectrics from a chemical bonding point of view.

cond-mat.mtrl-sci↗

arXiv2016

Effects of biaxial strain on the improper multiferroicity in h-LuFeO3 films

Elastic strain is potentially an important approach in tuning the properties of the improperly multiferroic hexagonal ferrites, the details of which have however been elusive due to the experimental difficulties. Employing the method of restrained thermal expansion, we have studied the effect of isothermal biaxial strain in the basal plane of h-LuFeO3 (001) films. The results indicate that a compressive biaxial strain significantly enhances the ferrodistortion, and the effect is larger at higher temperatures. The compressive biaxial strain and the enhanced ferrodistortion together, cause an increase in the electric polarization and a reduction in the canting of the weak ferromagnetic moments in h-LuFeO3, according to our first principle calculations. These findings are important for understanding the strain effect as well as the coupling between the lattice and the improper multiferroicity in h-LuFeO3. The experimental elucidation of the strain effect in h-LuFeO3 films also suggests that the restrained thermal expansion can be a viable method to unravel the strain effect in many other epitaxial thin film materials.

cond-mat.mtrl-sci↗

arXiv2015

SCAN: An Efficient Density Functional Yielding Accurate Structures and Energies of Diversely-Bonded Materials

Kohn-Sham density functional theory (DFT) is a widely-used electronic structure theory for materials as well as molecules. DFT is needed especially for large systems, ab initio molecular dynamics, and high-throughput searches for functional materials. DFT's accuracy and computational efficiency are limited by the approximation to its exchange-correlation energy. Currently, the local density approximation (LDA) and generalized gradient approximations (GGAs) dominate materials computation mainly due to their efficiency. We show here that the recently developed non-empirical strongly constrained and appropriately normed (SCAN) meta-GGA improves significantly over LDA and the standard Perdew-Burke-Ernzerhof GGA for geometries and energies of diversely-bonded materials (including covalent, metallic, ionic, hydrogen, and van der Waals bonds) at comparable efficiency. Thus SCAN may be useful even for soft matter. Often SCAN matches or improves upon the accuracy of a computationally expensive hybrid functional, at almost-GGA cost. SCAN is therefore expected to have a broad impact on materials science.

cond-mat.mtrl-sci↗

arXiv2011

Comparative Study of Structural and Electronic Properties of Cu-based Multinary Semiconductors

We present a systematic and comparative study of the structural and electronic properties of Cu-based ternary and quaternary semiconductors using first-principles electronic structure approaches. The important role that Cu d electrons play in determining their properties is illustrated by comparing results calculated with different exchange correlation energy functionals. We show that systematic improvement of the calculated anion displacement can be achieved by using the Heyd-Scuseria-Ernzerhof (HSE06) functional compared with the Perdew-Burke-Ernzerhof (PBE) functional. Quasiparticle band structures are then calculated within the G0W0 approximation using the crystal structures optimized within the HSE06 functional and starting from the PBE+U mean-field solution. Both the calculated quasiparticle band gaps and their systematic variation with chemical constituents agree very well with experiments. We also predict that the quasiparticle band gaps of the prototypical semiconductor Cu2ZnSnS4 in the kesterite (KS) phase is 1.65 eV and that of the stannite (ST) phase is 1.40 eV. These results are also consistent with available experimental values which vary from 1.45 to 1.6 eV.

cond-mat.mtrl-sci↗