arXiv subjects

Guorong Li

Publications and source records attributed to Guorong Li.

46 records · Page 3Linked to original sources

arXiv2020

Siamese Box Adaptive Network for Visual Tracking

Most of the existing trackers usually rely on either a multi-scale searching scheme or pre-defined anchor boxes to accurately estimate the scale and aspect ratio of a target. Unfortunately, they typically call for tedious and heuristic configurations. To address this issue, we propose a simple yet effective visual tracking framework (named Siamese Box Adaptive Network, SiamBAN) by exploiting the expressive power of the fully convolutional network (FCN). SiamBAN views the visual tracking problem as a parallel classification and regression problem, and thus directly classifies objects and regresses their bounding boxes in a unified FCN. The no-prior box design avoids hyper-parameters associated with the candidate boxes, making SiamBAN more flexible and general. Extensive experiments on visual tracking benchmarks including VOT2018, VOT2019, OTB100, NFS, UAV123, and LaSOT demonstrate that SiamBAN achieves state-of-the-art performance and runs at 40 FPS, confirming its effectiveness and efficiency. The code will be available at https://github.com/hqucv/siamban.

cs.CV↗

arXiv2019

Real-time Visual Object Tracking with Natural Language Description

In recent years, deep-learning-based visual object trackers have been studied thoroughly, but handling occlusions and/or rapid motion of the target remains challenging. In this work, we argue that conditioning on the natural language (NL) description of a target provides information for longer-term invariance, and thus helps cope with typical tracking challenges. However, deriving a formulation to combine the strengths of appearance-based tracking with the language modality is not straightforward. We propose a novel deep tracking-by-detection formulation that can take advantage of NL descriptions. Regions that are related to the given NL description are generated by a proposal network during the detection phase of the tracker. Our LSTM based tracker then predicts the update of the target from regions proposed by the NL based detection phase. In benchmarks, our method is competitive with state of the art trackers, while it outperforms all other trackers on targets with unambiguous and precise language annotations. It also beats the state-of-the-art NL tracker when initializing without a bounding box. Our method runs at over 30 fps on a single GPU.

cs.CV↗

arXiv2019

Spatiotemporal CNN for Video Object Segmentation

In this paper, we present a unified, end-to-end trainable spatiotemporal CNN model for VOS, which consists of two branches, i.e., the temporal coherence branch and the spatial segmentation branch. Specifically, the temporal coherence branch pretrained in an adversarial fashion from unlabeled video data, is designed to capture the dynamic appearance and motion cues of video sequences to guide object segmentation. The spatial segmentation branch focuses on segmenting objects accurately based on the learned appearance and motion cues. To obtain accurate segmentation results, we design a coarse-to-fine process to sequentially apply a designed attention module on multi-scale feature maps, and concatenate them to produce the final prediction. In this way, the spatial segmentation branch is enforced to gradually concentrate on object regions. These two branches are jointly fine-tuned on video segmentation sequences in an end-to-end manner. Several experiments are carried out on three challenging datasets (i.e., DAVIS-2016, DAVIS-2017 and Youtube-Object) to show that our method achieves favorable performance against the state-of-the-arts. Code is available at https://github.com/longyin880815/STCNN.

cs.CV↗

arXiv2018

Oxygen vacancies induced ferroelectricity in relaxors with ABO3 structure

The consideration of oxygen vacancies influence on the relaxors with perovskite structure was considered in the framework of Landau-Ginzburg-Devonshire phenomenological theory. The theory applicability for relaxors is based on the existence of some hidden soft phonon polar mode in them, and its frequency could be zero at some negative temperature TC*. Main attention was paid to PZN-PLZT relaxor described by formula 0.3Pb(Zn1/3Nb2/3)O3-0.7(Pb0.96La0.04(ZrxTi1-x)0.99O3) with x = 0.52, where earlier experimental investigation of oxygen vacancies influence on the polar properties was performed and the evidence of oxygen vacancies induced ferroelectricity was obtained. Since the oxygen vacancies are known to be elastic dipoles, they influence upon elastic and electric fields due to Vegard and flexoelectric coupling. We include the vacancies elastic and electrostrictive contribution into free energy functional. The calculations of the vacancies impact on polar properties were performed using their concentration distribution function. It was shown that the negative Curie temperature of a relaxor TC* is renormalized by the elastic dipoles due to the electrostrictive coupling and can become positive at some large enough concentration of the vacancies. We calculated the local polarization and electric field induced by the flexo-chemical coupling in dependence on the concentration of oxygen vacancies. The coexistence of FE phase and relaxor state can take place because of inhomogeneity of vacancies concentration distribution.

cond-mat.mtrl-sci↗

arXiv2018

The Unmanned Aerial Vehicle Benchmark: Object Detection and Tracking

With the advantage of high mobility, Unmanned Aerial Vehicles (UAVs) are used to fuel numerous important applications in computer vision, delivering more efficiency and convenience than surveillance cameras with fixed camera angle, scale and view. However, very limited UAV datasets are proposed, and they focus only on a specific task such as visual tracking or object detection in relatively constrained scenarios. Consequently, it is of great importance to develop an unconstrained UAV benchmark to boost related researches. In this paper, we construct a new UAV benchmark focusing on complex scenarios with new level challenges. Selected from 10 hours raw videos, about 80,000 representative frames are fully annotated with bounding boxes as well as up to 14 kinds of attributes (e.g., weather condition, flying altitude, camera view, vehicle category, and occlusion) for three fundamental computer vision tasks: object detection, single object tracking, and multiple object tracking. Then, a detailed quantitative study is performed using most recent state-of-the-art algorithms for each task. Experimental results show that the current state-of-the-art methods perform relative worse on our dataset, due to the new challenges appeared in UAV based real scenes, e.g., high density, small object, and camera motion. To our knowledge, our work is the first time to explore such issues in unconstrained scenes comprehensively.

cs.CV↗

arXiv2016

Hidden Phases Revealed at the Surface of Double-Layered Sr3(Ru1-xMnx)2O7

Double-layered Sr3Ru2O7 has received phenomenal consideration because it exhibits a plethora of exotic phases when perturbed. New phases emerge with the application of pressure, magnetic field, or doping. Here we show that creating a surface is an alternative and effective way to reveal hidden phases that are different from those seen in the bulk by investigating the surface properties of Sr3(Ru1-xMnx)2O7. Driven by the tilt distortion of RuO6 octahedra, the surface of Sr3Ru2O7 is less metallic than the bulk. In contrast, because of the vanishing of tilt and enhanced rotation with Mn-doping, the surface of Sr3(Ru0.84Mn0.16)2O7 is metallic while the bulk is insulating. Our result demonstrates that the electronic and structural properties at the surface are intimately coupled and consistent with quasi two-dimensional character.

cond-mat.str-el↗

arXiv2016

Interrogating the superconductor Ca10(Pt4As8)(Fe2-xPtxAs2)5 Layer-by-layer

Ever since the discovery of high-Tc superconductivity in layered cuprates, the roles that individual layers play have been debated, due to difficulty in layer-by-layer characterization. While there is similar challenge in many Fe-based layered superconductors, the newly-discovered Ca10(Pt4As8)(Fe2As2)5 provides opportunities to explore superconductivity layer by layer, because it contains both superconducting building blocks (Fe2As2 layers) and intermediate Pt4As8 layers. Cleaving a single crystal under ultra-high vacuum results in multiple terminations: an ordered Pt4As8 layer, two reconstructed Ca layers on the top of a Pt4As8 layer, and disordered Ca layer on the top of Fe2As2 layer. The electronic properties of individual layers are studied using scanning tunneling microscopy/spectroscopy (STM/S), which reveals different spectra for each surface. Remarkably superconducting coherence peaks are seen only on the ordered Ca/Pt4As8 layer. Our results indicate that an ordered structure with proper charge balance is required in order to preserve superconductivity.

cond-mat.supr-con↗

arXiv2013

Large and temperature-independent piezoelectric response in Pb(Mg1/3Nb2/3)O3-BaTiO3-PbTiO3

The temperature dependence of elastic, dielectric, and piezoelectric properties of (65-x)Pb(Mg1/3Nb2/3)O3-xBaTiO335-PbTiO3 ceramics with x=0, 1, 2, 3, and 4 was investigated. Compound with x=2 was found to exhibit a large piezoelectric response (d31=-170 pC/N, d33=530 pC/N at 300 K). Particularly, its d31 value was nearly a constant over a temperature range from 185 to 360 K. A broad ferroelectric phase transition tuned by BaTiO3 doping was deduced from the dielectric constant, elastic compliance constant and Raman spectra. The temperature-stable piezoelectric response was attributed to the counter-balance of contributions from the dielectric and elastic responses.

cond-mat.mtrl-sci↗

arXiv2010

BaFe2As2 Surface Domains and Domain Walls: Mirroring the Bulk Spin Structure

High-resolution scanning tunneling microscopy (STM) measurements on BaFe2As2-one of the parent compounds of the iron-based superconductors-reveals a (1x1) As-terminated unit cell on the (001) surface. However, there are significant differences of the surface unit cell compared to the bulk: only one of the two As atoms in the unit cell is imaged and domain walls between different (1x1) regions display a C2 symmetry at the surface. It should have been C2v if the STM image reflected the geometric structure of the surface or the orthorhombic bulk. The inequivalent As atoms and the bias dependence of the domain walls indicate that the origin of the STM image is primarily electronic not geometric. We argue that the surface electronic topography mirrors the bulk spin structure of BaFe2As2, via strong orbital-spin coupling.

cond-mat.supr-con↗

arXiv2009

Surface Geometric and Electronic Structure of BaFe2As2(001)

BaFe2As2 exhibits properties characteristic of the parent compounds of the newly discovered iron (Fe)-based high-TC superconductors. By combining the real space imaging of scanning tunneling microscopy/spectroscopy (STM/S) with momentum space quantitative Low Energy Electron Diffraction (LEED) we have identified the surface plane of cleaved BaFe2As2 crystals as the As terminated Fe-As layer - the plane where superconductivity occurs. LEED and STM/S data on the BaFe2As2(001) surface indicate an ordered arsenic (As) - terminated metallic surface without reconstruction or lattice distortion. It is surprising that the STM images the different Fe-As orbitals associated with the orthorhombic structure, not the As atoms in the surface plane.

cond-mat.mtrl-sci↗