arXiv subjects

Jianan Li

Publications and source records attributed to Jianan Li.

At least 91 records · Page 5Linked to original sources

arXiv2022

Electron Pairing and Nematicity in LaAlO$_3$/SrTiO$_3$ Nanostructures

Strongly correlated electronic systems exhibit a wealth of unconventional behavior stemming from strong electron-electron interactions. The LaAlO$_3$/SrTiO$_3$ (LAO/STO) heterostructure supports rich and varied low-temperature transport characteristics including low-density superconductivity, and electron pairing without superconductivity for which the microscopic origins is still not understood. LAO/STO also exhibits inexplicable signatures of electronic nematicity via nonlinear and anomalous Hall effects. Nanoscale control over the conductivity of the LAO/STO interface enables mesoscopic experiments that can probe these effects and address their microscopic origins. Here we report a direct correlation between electron pairing without superconductivity, anomalous Hall effect and electronic nematicity in quasi-1D ballistic nanoscale LAO/STO Hall crosses. The characteristic magnetic field at which the Hall coefficient changes directly coincides with the depairing of non-superconducting pairs showing a strong correlation between the two distinct phenomena. Angle-dependent Hall measurements further reveal an onset of electronic nematicity that again coincides with the electron pairing transition, unveiling a rotational symmetry breaking due to the transition from paired to unpaired phases at the interface. The results presented here highlights the influence of preformed electron pairs on the transport properties of LAO/STO and provide evidence of the elusive pairing ''glue'' that gives rise to electron pairing in SrTiO$_3$-based systems.

cond-mat.mes-hall↗

arXiv2022

Automated Optical Inspection of FAST's Reflector Surface using Drones and Computer Vision

The Five-hundred-meter Aperture Spherical radio Telescope (FAST) is the world's largest single-dish radio telescope. Its large reflecting surface achieves unprecedented sensitivity but is prone to damage, such as dents and holes, caused by naturally-occurring falling objects. Hence, the timely and accurate detection of surface defects is crucial for FAST's stable operation. Conventional manual inspection involves human inspectors climbing up and examining the large surface visually, a time-consuming and potentially unreliable process. To accelerate the inspection process and increase its accuracy, this work makes the first step towards automating the inspection of FAST by integrating deep-learning techniques with drone technology. First, a drone flies over the surface along a predetermined route. Since surface defects significantly vary in scale and show high inter-class similarity, directly applying existing deep detectors to detect defects on the drone imagery is highly prone to missing and misidentifying defects. As a remedy, we introduce cross-fusion, a dedicated plug-in operation for deep detectors that enables the adaptive fusion of multi-level features in a point-wise selective fashion, depending on local defect patterns. Consequently, strong semantics and fine-grained details are dynamically fused at different positions to support the accurate detection of defects of various scales and types. Our AI-powered drone-based automated inspection is time-efficient, reliable, and has good accessibility, which guarantees the long-term and stable operation of FAST.

cs.CV↗

arXiv2022

Metal-Enriched Neutral Gas Reservoir around a Strongly-lensed, Low-mass Galaxy at $z=4$ Identified by JWST/NIRISS and VLT/MUSE

Direct observations of low-mass, low-metallicity galaxies at $z\gtrsim4$ provide an indispensable opportunity for detailed inspection of the ionization radiation, gas flow, and metal enrichment in sources similar to those that reionized the Universe. Combining the James Webb Space Telescope (JWST), VLT/MUSE, and ALMA, we present detailed observations of a strongly lensed, low-mass ($\approx 10^{7.6}$ ${\rm M}_\odot$) galaxy at $z=3.98$ (also see Vanzella et al. 2022). We identify strong narrow nebular emission, including CIV $λ\lambda1548,1550$, HeII $\lambda1640$, OIII] $λ\lambda1661,1666$, [NeIII] $\lambda3868$, [OII] $\lambda3727$, and Balmer series of Hydrogen from this galaxy, indicating a metal-poor HII region ($\lesssim 0.12\ {\rm Z}_\odot$) powered by massive stars. Further, we detect a metal-enriched damped Ly$α$ system (DLA) associated with the galaxy with the HI column density of $N_{\rm{HI}}\approx 10^{21.8}$ cm$^{-2}$. The metallicity of the associated DLA may reach the super solar metallicity (${\gtrsim Z}_\odot$). Moreover, thanks to JWST and gravitational lensing, we present the resolved UV slope ($β$) map at the spatial resolution of $\approx 100$ pc at $z=4$, with steep UV slopes reaching $β\approx -2.5$ around three star-forming clumps. Combining with low-redshift analogs, our observations suggest that low-mass, low-metallicity galaxies, which dominate reionization, could be surrounded by a high covering fraction of the metal-enriched, neutral-gaseous clouds. This implies that the metal enrichment of low-mass galaxies is highly efficient, and further support that in low-mass galaxies, only a small fraction of ionizing radiation can escape through the interstellar or circumgalactic channels with low column-density neutral gas.

astro-ph.GA↗

arXiv2022

The Identification of a Dusty Multiarm Spiral Galaxy at $z=3.06$ with JWST and ALMA

Spiral arms serve crucial purposes in star formation and galaxy evolution. In this paper, we report the identification of A2744-DSG-$z3$, a dusty, multiarm spiral galaxy at $z=3.059$ using the James Webb Space Telescope (JWST) NIRISS imaging and grism spectroscopy. A2744-DSG-$z3$ was discovered as a gravitationally lensed sub-millimeter galaxy with ALMA. This is the most distant stellar spiral structure seen thus far, consistent with cosmological simulations which suggest $z\approx3$ as the epoch when spirals emerge. Thanks to the gravitational lensing and excellent spatial resolution of JWST, the spiral arms are resolved with a spatial resolution of $\approx290$\,pc. Based on SED fitting, the spiral galaxy has a de-lensed star formation rate of $85\pm30 \ M_{\odot}$ yr$^{-1}$, and a stellar mass of $\approx10^{10.6}\ M_{\odot}$, indicating that A2744-DSG-$z3$ is a main-sequence galaxy. After fitting the spiral arms, we find a stellar effective radius ($R_{e, \rm{star}}$) of $5.0\pm1.5$ kpc. Combing with ALMA measurements, we find that the effective radii ratio between dust and stars is $\approx0.4$, similar to {those} of massive SFGs at $z\sim2$, indicating a compact dusty core in A2744-DSG-$z3$. Moreover, this galaxy appears to be living in a group environment: including A2744-DSG-$z3$, at least three galaxies at $z=3.05 - 3.06$ {are} spectroscopically confirmed by JWST/NIRISS and ALMA, residing within a lensing-corrected projected scale of $\approx 70$ kpc. This, along with the asymmetric brightness profile, further suggests that the spiral arms may be triggered by minor merger events at $z\gtrsim3$.

astro-ph.GA↗

arXiv2022

RescureService: A Benchmark Microservice System for the Research of Mobile Edge and Cloud Computing

The dramatic development of cloud and edge computing allows for better Quality of Service (QoS) in many scenarios by deploying services on cloud and edge servers. Microservice technology is also adopted in these scenarios to decompose complex business logic into many small independent services. Meanwhile, as microservice systems continue to grow, providing stable QoS in these systems becomes a challenge, and many different approaches have been proposed for stable QoS. However, the microservice systems used in the experiments of these work have problems such as the low number of services and a single type of service. Therefore, we developed the open-source benchmark microservice system RescureService with 20+ services, including database, front-end, business logic, data processing, and artificial intelligence services in the disaster relief scenario. Measuring tools are provided to measure the service properties to help researchers prepare experimental data, and the service properties pre-measured are also presented. Meanwhile, the fulfillment of benchmark requirements is detailed, and the results show that our RescureService meets the requirements of a benchmark system in research. Moreover, instructions are given to describe adopting our system in service computing as examples.

cs.DC↗

arXiv2022

CAGroup3D: Class-Aware Grouping for 3D Object Detection on Point Clouds

We present a novel two-stage fully sparse convolutional 3D object detection framework, named CAGroup3D. Our proposed method first generates some high-quality 3D proposals by leveraging the class-aware local group strategy on the object surface voxels with the same semantic predictions, which considers semantic consistency and diverse locality abandoned in previous bottom-up approaches. Then, to recover the features of missed voxels due to incorrect voxel-wise segmentation, we build a fully sparse convolutional RoI pooling module to directly aggregate fine-grained spatial information from backbone for further proposal refinement. It is memory-and-computation efficient and can better encode the geometry-specific features of each 3D proposal. Our model achieves state-of-the-art 3D detection performance with remarkable gains of +\textit{3.6\%} on ScanNet V2 and +\textit{2.6}\% on SUN RGB-D in term of mAP@0.25. Code will be available at https://github.com/Haiyang-W/CAGroup3D.

cs.CV↗

arXiv2022

FusionRCNN: LiDAR-Camera Fusion for Two-stage 3D Object Detection

3D object detection with multi-sensors is essential for an accurate and reliable perception system of autonomous driving and robotics. Existing 3D detectors significantly improve the accuracy by adopting a two-stage paradigm which merely relies on LiDAR point clouds for 3D proposal refinement. Though impressive, the sparsity of point clouds, especially for the points far away, making it difficult for the LiDAR-only refinement module to accurately recognize and locate objects.To address this problem, we propose a novel multi-modality two-stage approach named FusionRCNN, which effectively and efficiently fuses point clouds and camera images in the Regions of Interest(RoI). FusionRCNN adaptively integrates both sparse geometry information from LiDAR and dense texture information from camera in a unified attention mechanism. Specifically, it first utilizes RoIPooling to obtain an image set with a unified size and gets the point set by sampling raw points within proposals in the RoI extraction step; then leverages an intra-modality self-attention to enhance the domain-specific features, following by a well-designed cross-attention to fuse the information from two modalities.FusionRCNN is fundamentally plug-and-play and supports different one-stage methods with almost no architectural changes. Extensive experiments on KITTI and Waymo benchmarks demonstrate that our method significantly boosts the performances of popular detectors.Remarkably, FusionRCNN significantly improves the strong SECOND baseline by 6.14% mAP on Waymo, and outperforms competing two-stage approaches. Code will be released soon at https://github.com/xxlbigbrother/Fusion-RCNN.

cs.CV↗

arXiv2022

Spatially resolved molecular interstellar medium in a $z=6.6$ quasar host galaxy

We present high spatial resolution (0.4", 2.2kpc) observations of the CO(6-5), CO(7-6) and [CI] lines and dust continuum emission from the interstellar medium in the host galaxy of the quasar J0305$-$3150 at $z=6.6$. These, together with archival [CII] data at comparable spatial resolution, enable studies of the spatial distribution and kinematics between the ISM in different phases. When comparing the radial profiles of CO, [CII] and the dust continuum, we find that the CO and dust continuum exhibit similar spatial distributions, both of which are less extended than the [CII], indicating that the CO and dust continuum are tracing the same gas component, while the [CII] is tracing a more extended one. In addition, we derive the radial profiles of the [CII]/CO, [CII]/far-infrared (FIR), CO/FIR, and dust continuum $S_{98.7 \rm GHz}/S_{258.1 \rm GHz}$ ratios. We find a decreasing $S_{98.7 \rm GHz}/S_{258.1 \rm GHz}$ ratio with radius, possibly indicating a decrease of dust optical depth with increasing radius. We also detect some of the ISM lines and continuum emission in the companion galaxies previously discovered in the field around J0305$-$3150. Through comparing the line-to-line and line-to-FIR ratios, we find no significant differences between the quasar and its companion galaxies.

astro-ph.GA↗

arXiv2022

Molecular gas in z~6 quasar host galaxies

We investigate the molecular gas content of z~6 quasar host galaxies using the IRAM / Northern Extended Millimeter Array. We target the 3mm dust continuum, and the line emission from CO(6-5), CO(7-6), [CI]2-1 in 10 infra-red-luminous quasars that have been previously studied in their 1mm dust continuum and [CII] line emission. We detect CO(7-6) at various degrees of significance in all the targeted sources, thus doubling the number of such detections in z~6 quasars. The 3mm to 1mm flux density ratios are consistent with a modified black body spectrum with a dust temperature $T_{dust}$~47 K and an optical depth $τ_ν$=0.2 at the [CII] frequency. Our study provides us with four independent ways to estimate the molecular gas mass, $M_{H2}$, in the targeted quasars. This allows us to set constraints on various parameters used in the derivation of molecular gas mass estimates, such as the mass per luminosity ratios $α_{CO}$ and $α_{[CII]}$, the gas-to-dust ratio $δ_{g/d}$, and the carbon abundance [C]/H2. Leveraging either on the dust, CO, [CI], or [CII] emission yields mass estimates of the entire sample in the range $M_{H2}$~$10^{10}$ to $10^{11}$ M$_{\odot}$. We compare the observed luminosities of dust, [CII], [CI], and CO(7-6) with predictions from photo-dissociation and X-ray dominated regions. We find that the former provide better model fits to our data, assuming that the bulk of the emission arises from dense ($n_H>10^4$ cm$^{-3}$) clouds with a column density $N_{H}$~$10^{23}$ cm$^{-2}$, exposed to a radiation field with intensity $G_0$~$10^3$ (in Habing units). Our analysis reiterates the presence of massive reservoirs of molecular gas fueling star formation and nuclear accretion in $z$~6 quasar host galaxies. It also highlights the power of combined 3mm and 1mm observations for quantitative studies of the dense gas content in massive galaxies at cosmic dawn.

astro-ph.GA↗

arXiv2022

Exploring the radio spectral energy distribution of the ultraluminous radio-quiet quasar SDSS J0100+2802 at redshift 6.3

We report deep Karl G. Jansky Very Large Array (VLA) observations of the optically ultraluminous and radio-quiet quasar SDSS J010013.02 + 280225.8 (hereafter J0100+2802) at redshift $z=$6.3. We detected the radio continuum emission at 1.5 GHz, 6 GHz, and 10 GHz. This leads to a radio power-law spectral index of $α= -0.52\pm0.18$ ($S \propto ν^α$). The radio source is unresolved in all VLA bands with an upper limit to the size of $0.2^{\prime \prime}$ (i.e., $\sim$ 1.1 kpc) at 10 GHz. We find variability in the flux density (increase by $\sim 33\%$) and the spectral index (steepened) between observations in 2016 and 2017. We also find that the VLA 1.5 GHz flux density observed in the same year is 1.5 times that detected with the Very Long Baseline Array (VLBA) in 2016 at the same frequency. This difference suggests that half of the radio emission from J0100+2802 comes from a compact core within 40 pc, and the rest comes from the surrounding few kpc area which is diffuse and resolved out in the VLBA observations. The diffuse emission is four times brighter than that would be expected if driven by star formation. We conclude that the central active galactic nucleus is the dominant power engine of the radio emission in J0100+2802.

astro-ph.GA↗

arXiv2022

RTNet: Relation Transformer Network for Diabetic Retinopathy Multi-lesion Segmentation

Automatic diabetic retinopathy (DR) lesions segmentation makes great sense of assisting ophthalmologists in diagnosis. Although many researches have been conducted on this task, most prior works paid too much attention to the designs of networks instead of considering the pathological association for lesions. Through investigating the pathogenic causes of DR lesions in advance, we found that certain lesions are closed to specific vessels and present relative patterns to each other. Motivated by the observation, we propose a relation transformer block (RTB) to incorporate attention mechanisms at two main levels: a self-attention transformer exploits global dependencies among lesion features, while a cross-attention transformer allows interactions between lesion and vessel features by integrating valuable vascular information to alleviate ambiguity in lesion detection caused by complex fundus structures. In addition, to capture the small lesion patterns first, we propose a global transformer block (GTB) which preserves detailed information in deep network. By integrating the above blocks of dual-branches, our network segments the four kinds of lesions simultaneously. Comprehensive experiments on IDRiD and DDR datasets well demonstrate the superiority of our approach, which achieves competitive performance compared to state-of-the-arts.

eess.IV↗

arXiv2021

Delving into Sample Loss Curve to Embrace Noisy and Imbalanced Data

Corrupted labels and class imbalance are commonly encountered in practically collected training data, which easily leads to over-fitting of deep neural networks (DNNs). Existing approaches alleviate these issues by adopting a sample re-weighting strategy, which is to re-weight sample by designing weighting function. However, it is only applicable for training data containing only either one type of data biases. In practice, however, biased samples with corrupted labels and of tailed classes commonly co-exist in training data. How to handle them simultaneously is a key but under-explored problem. In this paper, we find that these two types of biased samples, though have similar transient loss, have distinguishable trend and characteristics in loss curves, which could provide valuable priors for sample weight assignment. Motivated by this, we delve into the loss curves and propose a novel probe-and-allocate training strategy: In the probing stage, we train the network on the whole biased training data without intervention, and record the loss curve of each sample as an additional attribute; In the allocating stage, we feed the resulting attribute to a newly designed curve-perception network, named CurveNet, to learn to identify the bias type of each sample and assign proper weights through meta-learning adaptively. The training speed of meta learning also blocks its application. To solve it, we propose a method named skip layer meta optimization (SLMO) to accelerate training speed by skipping the bottom layers. Extensive synthetic and real experiments well validate the proposed method, which achieves state-of-the-art performance on multiple challenging benchmarks.

cs.LG↗

arXiv2021

PAPooling: Graph-based Position Adaptive Aggregation of Local Geometry in Point Clouds

Fine-grained geometry, captured by aggregation of point features in local regions, is crucial for object recognition and scene understanding in point clouds. Nevertheless, existing preeminent point cloud backbones usually incorporate max/average pooling for local feature aggregation, which largely ignores points' positional distribution, leading to inadequate assembling of fine-grained structures. To mitigate this bottleneck, we present an efficient alternative to max pooling, Position Adaptive Pooling (PAPooling), that explicitly models spatial relations among local points using a novel graph representation, and aggregates features in a position adaptive manner, enabling position-sensitive representation of aggregated features. Specifically, PAPooling consists of two key steps, Graph Construction and Feature Aggregation, respectively in charge of constructing a graph with edges linking the center point with every neighboring point in a local region to map their relative positional information to channel-wise attentive weights, and adaptively aggregating local point features based on the generated weights through Graph Convolution Network (GCN). PAPooling is simple yet effective, and flexible enough to be ready to use for different popular backbones like PointNet++ and DGCNN, as a plug-andplay operator. Extensive experiments on various tasks ranging from 3D shape classification, part segmentation to scene segmentation well demonstrate that PAPooling can significantly improve predictive accuracy, while with minimal extra computational overhead. Code will be released.

cs.CV↗

arXiv2021

A global quadratic speed-up for computing the principal eigenvalue of Perron-like operators

We consider a new algorithm in light of the min-max Collatz-Wielandt formalism to compute the principal eigenvalue and the eigenvector (eigen-function) for a class of positive Perron-Frobenius-like operators. Such operators are natural generalizations of the usual nonnegative primitive matrices. These have nontrivial applications in PDE problems such as computing the principal eigenvalue of Dirichlet Laplacian operators on general domains. We rigorously prove that for general initial data the corresponding numerical iterates converge globally to the unique principal eigenvalue with quadratic convergence. We show that the quadratic convergence is sharp with compatible upper and lower bounds. We demonstrate the effectiveness of the scheme via several illustrative numerical examples.

math.NA↗

arXiv2021

Massive molecular outflow and 100 kpc extended cold halo gas in the enormous Ly$α$ nebula of QSO 1228+3128

The link between the circum-galactic medium (CGM) and the stellar growth of massive galaxies at high-$z$ depends on the properties of the widespread cold molecular gas. As part of the SUPERCOLD-CGM survey (Survey of Protocluster ELANe Revealing CO/\ci\ in the Ly$α$-Detected CGM), we present the radio-loud QSO Q1228+3128 at $z=2.2218$, which is embedded in an enormous Ly$α$ nebula. ALMA+ACA observations of CO(4-3) reveal both a massive molecular outflow, and a more extended molecular gas reservoir across $\sim$100 kpc in the CGM each containing a mass of M$_{\rm H2}$\,$\sim$\,4$-$5\,$\times$\,10$^{10}$ M$_{\odot}$. The outflow and molecular CGM are aligned spatially, along the direction of an inner radio jet. After re-analysis of Ly$α$ data of Q1228+3128 from the Keck Cosmic Web Imager, we found that the velocity of the extended CO agrees with the redshift derived from the Ly$α$ nebula and the bulk velocity of the massive outflow. We propose a scenario where the radio source in Q1228+3128 is driving the molecular outflow and perhaps also enriching or cooling the CGM. In addition, we found that the extended CO emission is nearly perpendicular to the extended Ly$α$ nebula spatially, indicating that the two gas phases are not well mixed, and possibly even represent different phenomena (e.g., outflow vs. infall). Our results provide crucial evidence in support of predicted baryonic recycling processes that drive the early evolution of massive galaxies.

astro-ph.GA↗

arXiv2021

Pyramid Correlation based Deep Hough Voting for Visual Object Tracking

Most of the existing Siamese-based trackers treat tracking problem as a parallel task of classification and regression. However, some studies show that the sibling head structure could lead to suboptimal solutions during the network training. Through experiments we find that, without regression, the performance could be equally promising as long as we delicately design the network to suit the training objective. We introduce a novel voting-based classification-only tracking algorithm named Pyramid Correlation based Deep Hough Voting (short for PCDHV), to jointly locate the top-left and bottom-right corners of the target. Specifically we innovatively construct a Pyramid Correlation module to equip the embedded feature with fine-grained local structures and global spatial contexts; The elaborately designed Deep Hough Voting module further take over, integrating long-range dependencies of pixels to perceive corners; In addition, the prevalent discretization gap is simply yet effectively alleviated by increasing the spatial resolution of the feature maps while exploiting channel-space relationships. The algorithm is general, robust and simple. We demonstrate the effectiveness of the module through a series of ablation experiments. Without bells and whistles, our tracker achieves better or comparable performance to the SOTA algorithms on three challenging benchmarks (TrackingNet, GOT-10k and LaSOT) while running at a real-time speed of 80 FPS. Codes and models will be released.

cs.CV↗

arXiv2021

The 2nd Anti-UAV Workshop & Challenge: Methods and Results

The 2nd Anti-UAV Workshop \& Challenge aims to encourage research in developing novel and accurate methods for multi-scale object tracking. The Anti-UAV dataset used for the Anti-UAV Challenge has been publicly released. There are two subsets in the dataset, $i.e.$, the test-dev subset and test-challenge subset. Both subsets consist of 140 thermal infrared video sequences, spanning multiple occurrences of multi-scale UAVs. Around 24 participating teams from the globe competed in the 2nd Anti-UAV Challenge. In this paper, we provide a brief summary of the 2nd Anti-UAV Workshop \& Challenge including brief introductions to the top three methods.The submission leaderboard will be reopened for researchers that are interested in the Anti-UAV challenge. The benchmark dataset and other information can be found at: https://anti-uav.github.io/.

cs.CV↗

arXiv2021

Updatable Siamese Tracker with Two-stage One-shot Learning

Offline Siamese networks have achieved very promising tracking performance, especially in accuracy and efficiency. However, they often fail to track an object in complex scenes due to the incapacity in online update. Traditional updaters are difficult to process the irregular variations and sampling noises of objects, so it is quite risky to adopt them to update Siamese networks. In this paper, we first present a two-stage one-shot learner, which can predict the local parameters of primary classifier with object samples from diverse stages. Then, an updatable Siamese network is proposed based on the learner (SiamTOL), which is able to complement online update by itself. Concretely, we introduce an extra inputting branch to sequentially capture the latest object features, and design a residual module to update the initial exemplar using these features. Besides, an effective multi-aspect training loss is designed for our network to avoid overfit. Extensive experimental results on several popular benchmarks including OTB100, VOT2018, VOT2019, LaSOT, UAV123 and GOT10k manifest that the proposed tracker achieves the leading performance and outperforms other state-of-the-art methods

cs.CV↗