arXiv subjects

Yu Rong

Publications and source records attributed to Yu Rong.

At least 163 records · Page 9Linked to original sources

arXiv2020

Chasing the Tail in Monocular 3D Human Reconstruction with Prototype Memory

Deep neural networks have achieved great progress in single-image 3D human reconstruction. However, existing methods still fall short in predicting rare poses. The reason is that most of the current models perform regression based on a single human prototype, which is similar to common poses while far from the rare poses. In this work, we 1) identify and analyze this learning obstacle and 2) propose a prototype memory-augmented network, PM-Net, that effectively improves performances of predicting rare poses. The core of our framework is a memory module that learns and stores a set of 3D human prototypes capturing local distributions for either common poses or rare poses. With this formulation, the regression starts from a better initialization, which is relatively easier to converge. Extensive experiments on several widely employed datasets demonstrate the proposed framework's effectiveness compared to other state-of-the-art methods. Notably, our approach significantly improves the models' performances on rare poses while generating comparable results on other samples.

cs.CV↗

arXiv2020

Deep Multimodal Fusion by Channel Exchanging

Deep multimodal fusion by using multiple sources of data for classification or regression has exhibited a clear advantage over the unimodal counterpart on various applications. Yet, current methods including aggregation-based and alignment-based fusion are still inadequate in balancing the trade-off between inter-modal fusion and intra-modal processing, incurring a bottleneck of performance improvement. To this end, this paper proposes Channel-Exchanging-Network (CEN), a parameter-free multimodal fusion framework that dynamically exchanges channels between sub-networks of different modalities. Specifically, the channel exchanging process is self-guided by individual channel importance that is measured by the magnitude of Batch-Normalization (BN) scaling factor during training. The validity of such exchanging process is also guaranteed by sharing convolutional filters yet keeping separate BN layers across modalities, which, as an add-on benefit, allows our multimodal architecture to be almost as compact as a unimodal network. Extensive experiments on semantic segmentation via RGB-D data and image translation through multi-domain input verify the effectiveness of our CEN compared to current state-of-the-art methods. Detailed ablation studies have also been carried out, which provably affirm the advantage of each component we propose. Our code is available at https://github.com/yikaiw/CEN.

cs.CV↗

arXiv2020

Non-Contact Vital Signs Detection with UAV-Borne Radars

Airborne radar carried on-board unmanned aerial vehicles (UAV) is serving as the harbinger of new remote sensing applications for security and rescue in inclement environments. The mobility and agility of UAVs along with intelligent on-board sensors (cameras, acoustics, and radar) are more effective during the early stages of disaster response. The ability of radars to penetrate through objects and operate during low visibility conditions enables detection of occluded human subjects on and under debris when other sensing modalities fail. Recently, radars have been deployed on UAVs to measure minute human physiological parameters such as respiratory and heart rates while sensing through clothing and building materials. Aggregating radar measurements with the information from other sensors is broadening the applications of drones in life-critical situations. Signal processing techniques are critical in enabling UAV-borne radars for human vital sign detection (VSD) in multiple operation modes. Novel radar configurations such as in a UAV swarm and tethered UAVs are required to facilitate multi-tasking and high endurance, respectively. This paper provides an overview of recent advances in UAV-borne VSD with a focus on deployment modes and processing methods.

eess.SP↗

arXiv2020

Self-Supervised Graph Transformer on Large-Scale Molecular Data

How to obtain informative representations of molecules is a crucial prerequisite in AI-driven drug design and discovery. Recent researches abstract molecules as graphs and employ Graph Neural Networks (GNNs) for molecular representation learning. Nevertheless, two issues impede the usage of GNNs in real scenarios: (1) insufficient labeled molecules for supervised training; (2) poor generalization capability to new-synthesized molecules. To address them both, we propose a novel framework, GROVER, which stands for Graph Representation frOm self-superVised mEssage passing tRansformer. With carefully designed self-supervised tasks in node-, edge- and graph-level, GROVER can learn rich structural and semantic information of molecules from enormous unlabelled molecular data. Rather, to encode such complex information, GROVER integrates Message Passing Networks into the Transformer-style architecture to deliver a class of more expressive encoders of molecules. The flexibility of GROVER allows it to be trained efficiently on large-scale molecular dataset without requiring any supervision, thus being immunized to the two issues mentioned above. We pre-train GROVER with 100 million parameters on 10 million unlabelled molecules -- the biggest GNN and the largest training dataset in molecular representation learning. We then leverage the pre-trained GROVER for molecular property prediction followed by task-specific fine-tuning, where we observe a huge improvement (more than 6% on average) from current state-of-the-art methods on 11 challenging benchmarks. The insights we gained are that well-designed self-supervision losses and largely-expressive pre-trained models enjoy the significant potential on performance boosting.

q-bio.BM↗

arXiv2020

Cascade-BGNN: Toward Efficient Self-supervised Representation Learning on Large-scale Bipartite Graphs

Bipartite graphs have been used to represent data relationships in many data-mining applications such as in E-commerce recommendation systems. Since learning in graph space is more complicated than in Euclidian space, recent studies have extensively utilized neural nets to effectively and efficiently embed a graph's nodes into a multidimensional space. However, this embedding method has not yet been applied to large-scale bipartite graphs. Existing techniques either cannot be scaled to large-scale bipartite graphs that have limited labels or cannot exploit the unique structure of bipartite graphs, which have distinct node features in two domains. Thus, we propose Cascade Bipartite Graph Neural Networks, Cascade-BGNN, a novel node representation learning for bipartite graphs that is domain-consistent, self-supervised, and efficient. To efficiently aggregate information both across and within the two partitions of a bipartite graph, BGNN utilizes a customized Inter-domain Message Passing (IDMP) and Intra-domain Alignment (IDA), which is our adaptation of adversarial learning, for message aggregation across and within partitions, respectively. BGNN is trained in a self-supervised manner. Moreover, we formulate a multi-layer BGNN in a cascaded training manner to enable multi-hop relationship modeling while improving training efficiency. Extensive experiments on several datasets of varying scales verify the effectiveness and efficiency of BGNN over baselines. Our design is further affirmed through theoretical analysis for domain alignment. The scalability of BGNN is additionally verified through its demonstrated rapid training speed and low memory cost over a large-scale real-world bipartite graph.

cs.SI↗

arXiv2020

Graph Information Bottleneck for Subgraph Recognition

Given the input graph and its label/property, several key problems of graph learning, such as finding interpretable subgraphs, graph denoising and graph compression, can be attributed to the fundamental problem of recognizing a subgraph of the original one. This subgraph shall be as informative as possible, yet contains less redundant and noisy structure. This problem setting is closely related to the well-known information bottleneck (IB) principle, which, however, has less been studied for the irregular graph data and graph neural networks (GNNs). In this paper, we propose a framework of Graph Information Bottleneck (GIB) for the subgraph recognition problem in deep graph learning. Under this framework, one can recognize the maximally informative yet compressive subgraph, named IB-subgraph. However, the GIB objective is notoriously hard to optimize, mostly due to the intractability of the mutual information of irregular graph data and the unstable optimization process. In order to tackle these challenges, we propose: i) a GIB objective based-on a mutual information estimator for the irregular graph data; ii) a bi-level optimization scheme to maximize the GIB objective; iii) a connectivity loss to stabilize the optimization process. We evaluate the properties of the IB-subgraph in three application scenarios: improvement of graph classification, graph interpretation and graph denoising. Extensive experiments demonstrate that the information-theoretic IB-subgraph enjoys superior graph properties.

cs.LG↗

arXiv2020

FrankMocap: Fast Monocular 3D Hand and Body Motion Capture by Regression and Integration

Although the essential nuance of human motion is often conveyed as a combination of body movements and hand gestures, the existing monocular motion capture approaches mostly focus on either body motion capture only ignoring hand parts or hand motion capture only without considering body motion. In this paper, we present FrankMocap, a motion capture system that can estimate both 3D hand and body motion from in-the-wild monocular inputs with faster speed (9.5 fps) and better accuracy than previous work. Our method works in near real-time (9.5 fps) and produces 3D body and hand motion capture outputs as a unified parametric model structure. Our method aims to capture 3D body and hand motion simultaneously from challenging in-the-wild monocular videos. To construct FrankMocap, we build the state-of-the-art monocular 3D "hand" motion capture method by taking the hand part of the whole body parametric model (SMPL-X). Our 3D hand motion capture output can be efficiently integrated to monocular body motion capture output, producing whole body motion results in a unified parrametric model structure. We demonstrate the state-of-the-art performance of our hand motion capture system in public benchmarks, and show the high quality of our whole body motion capture result in various challenging real-world scenes, including a live demo scenario.

cs.CV↗

arXiv2020

The Blue Compact Dwarf Galaxy VCC 848 Formed by Dwarf-Dwarf Merging: HI Gas, Star Formation and Numerical Simulations

A clear link between a dwarf-dwarf merger event and enhanced star formation (SF) in the recent past was recently identified in the gas-dominated merger remnant VCC 848, offering by far the clearest view of a gas-rich late-stage dwarf-dwarf merger. We present a joint analysis of JVLA HI emission-line mapping, optical imaging and numerical simulations of VCC 848, in order to examine the impact of the merger on the stellar and gaseous distributions. VCC 848 has less than 30% of its HI gas concentrated within the central high-surface-brightness star-forming region, while the remaining HI is entrained in outlying tidal features. Particularly, a well-defined tidal arm reaches N(HI) comparable to the galaxy center but lacks SF. The molecular gas mass inferred from the current SF rate (SFR) dominates over the atomic gas mass in the central ~ 1.5 kpc. VCC 848 is consistent with being a main-sequence star-forming galaxy for its current stellar mass and SFR. The HII region luminosity distribution largely agrees with that of normal dwarf irregulars with similar luminosities, except that the brightest HII region is extraordinarily luminous. Our N-body/hydrodynamical simulations imply that VCC 848 is a merger between a gas-dominated primary progenitor and a gas-bearing star-dominated secondary. The progenitors had their first passage on a near-radial non-coplanar orbit more than 1 Gyr ago. The merger did not build up a core as compact as typical compact dwarfs with centralized starburst, which may be partly ascribed to the star-dominated nature of the secondary, and in a general sense, a negative stellar feedback following intense starbursts triggered at early stages of the merger.

astro-ph.GA↗

arXiv2020

Exploring the origin of ultra-diffuse galaxies in clusters from their primordial alignment

We find that the minor axes of the ultra-diffuse galaxies (UDGs) in Abell 2634 tend to be aligned with the major axis of the central dominant galaxy, at a $\gtrsim 95\%$ confidence level. This alignment is produced by the bright UDGs with the absolute magnitudes $M_r<-15.3$ mag, and outer-region UDGs with $R>0.5R_{200}$. The alignment signal implies that these bright, outer-region UDGs are very likely to acquire their angular momenta from the vortices around the large-scale filament before they were accreted into A2634, and form their extended stellar bodies outside of the cluster; in this scenario, the orientations of their primordial angular momenta, which are roughly shown by their minor axes on the images, should tend to be parallel to the elongation of the large-scale filament. When these UDGs fell into the unrelaxed cluster A2634 along the filament, they could still preserve their primordial alignment signal before violent relaxation and encounters. These bright, outer-region UDGs in A2634 are very unlikely to be the descendants of the high-surface-brightness dwarf progenitors under tidal interactions with the central dominant galaxy in the cluster environment. Our results indicate that the primordial alignment could be a useful probe of the origin of UDGs in large-scale structures.

astro-ph.GA↗

arXiv2020

Intrinsic Morphology of Ultra-diffuse Galaxies

With the published data of apparent axis ratios for 1109 ultra-diffuse galaxies (UDGs) located in 17 low-redshift (z~ 0.020 - 0.063) galaxy clusters and 84 UDGs in 2 intermediate-redshift (z~ 0.308 - 0.348) clusters, we take advantage of a Markov Chain Monte Carlo approach and assume a ubiquitous triaxial model to investigate the intrinsic morphologies of UDGs. In contrast to the conclusion of Burkert (2017), i.e., the underlying shapes of UDGs are purely prolate ($C=B<A$), we find that the data favor the oblate-triaxial models ($C<B\lesssim A$) over the nearly prolate ones. We also find that the intrinsic morphologies of UDGs are relevant to their stellar masses/luminosities, environments, and redshifts. First, for the low-redshift UDGs in the same environment, the more-luminous ones are always thicker than the less-luminous counterparts, possibly due to the more voilent internal supernovae feedback or external tidal interactions for the progenitors of the more-luminous UDGs. The UDG thickness dependence on luminosity is distinct from that of the typical quiescent dwarf ellipticals (dEs) and dwarf spheroidals (dSphs) in the local clusters and groups, but resembles that of massive galaxies; in this sense, UDGs may not be simply treated as an extension of the dE/dSph class with similar evolutionary histories. Second, for the low-redshift UDGs within the same luminosity range, the ones with smaller cluster-centric distances are more puffed-up, probably attributed to tidal interactions. Finally, the intermediate-redshift cluster UDGs are more flattened, which plausibly suggests a `disky' origin for high-redshift, initial UDGs.

astro-ph.GA↗

arXiv2020

Lessons on Star-forming Ultra-diffuse Galaxies from The Stacked Spectra of Sloan Digital Sky Survey

We investigate the on-average properties for 28 star-forming ultra-diffuse galaxies (UDGs) located in low-density environments, by stacking their spectra from the Sloan Digital Sky Survey. These relatively-isolated UDGs, with stellar masses of $\log_{10}(M_*/M_{\odot})\sim 8.57\pm0.29$, have the on-average total-stellar-metallicity [M/H]$\sim -0.82\pm0.14$, iron-metallicity [Fe/H]$\sim -1.00\pm0.16$, stellar age $t_*\sim5.2\pm0.5$ Gyr, $α$-enhancement [$α$/Fe]$\sim 0.24\pm0.10$, and oxygen abundance 12+log(O/H)$\sim 8.16\pm0.06$, as well as central stellar velocity dispersion $54\pm12$ km/s. On the star-formation rate versus stellar mass diagram, these UDGs are located lower than the extrapolated star-forming main sequence from the massive spirals, but roughly follow the main sequence of low-surface-brightness dwarf galaxies. We find that these star-forming UDGs are not particularly metal-poor or metal-rich for their stellar masses, as compared with the metallicity-mass relations of the nearby typical dwarfs. With the UDG data of this work and previous studies, we also find a coarse correlation between [Fe/H] and magnesium-element enhancement [Mg/Fe] for UDGs: [Mg/Fe]$\simeq-0.43(\pm0.26)$[Fe/H]$-0.14(\pm0.40)$.

astro-ph.GA↗

arXiv2020

Inverse Graph Identification: Can We Identify Node Labels Given Graph Labels?

Graph Identification (GI) has long been researched in graph learning and is essential in certain applications (e.g. social community detection). Specifically, GI requires to predict the label/score of a target graph given its collection of node features and edge connections. While this task is common, more complex cases arise in practice---we are supposed to do the inverse thing by, for example, grouping similar users in a social network given the labels of different communities. This triggers an interesting thought: can we identify nodes given the labels of the graphs they belong to? Therefore, this paper defines a novel problem dubbed Inverse Graph Identification (IGI), as opposed to GI. Upon a formal discussion of the variants of IGI, we choose a particular case study of node clustering by making use of the graph labels and node features, with an assistance of a hierarchical graph that further characterizes the connections between different graphs. To address this task, we propose Gaussian Mixture Graph Convolutional Network (GMGCN), a simple yet effective method that makes the node-level message passing process using Graph Attention Network (GAT) under the protocol of GI and then infers the category of each node via a Gaussian Mixture Layer (GML). The training of GMGCN is further boosted by a proposed consensus loss to take advantage of the structure of the hierarchical graph. Extensive experiments are conducted to test the rationality of the formulation of IGI. We verify the superiority of the proposed method compared to other baselines on several benchmarks we have built up. We will release our codes along with the benchmark data to facilitate more research attention to the IGI problem.

cs.LG↗

arXiv2020

Multi-View Graph Neural Networks for Molecular Property Prediction

The crux of molecular property prediction is to generate meaningful representations of the molecules. One promising route is to exploit the molecular graph structure through Graph Neural Networks (GNNs). It is well known that both atoms and bonds significantly affect the chemical properties of a molecule, so an expressive model shall be able to exploit both node (atom) and edge (bond) information simultaneously. Guided by this observation, we present Multi-View Graph Neural Network (MV-GNN), a multi-view message passing architecture to enable more accurate predictions of molecular properties. In MV-GNN, we introduce a shared self-attentive readout component and disagreement loss to stabilize the training process. This readout component also renders the whole architecture interpretable. We further boost the expressive power of MV-GNN by proposing a cross-dependent message passing scheme that enhances information communication of the two views, which results in the MV-GNN^cross variant. Lastly, we theoretically justify the expressiveness of the two proposed models in terms of distinguishing non-isomorphism graphs. Extensive experiments demonstrate that MV-GNN models achieve remarkably superior performance over the state-of-the-art models on a variety of challenging benchmarks. Meanwhile, visualization results of the node importance are consistent with prior knowledge, which confirms the interpretability power of MV-GNN models.

q-bio.QM↗

arXiv2020

The Next Generation Fornax Survey (NGFS): VII. A MUSE view of the nuclear star clusters in Fornax dwarf galaxies

Clues to the formation and evolution of Nuclear Star Clusters (NSCs) lie in their stellar populations. However, these structures are often very faint compared to their host galaxy, and spectroscopic analysis of NSCs is hampered by contamination of light from the rest of the system. With the introduction of wide-field IFU spectrographs, new techniques have been developed to model the light from different components within galaxies, making it possible to cleanly extract the spectra of the NSCs and study their properties with minimal contamination from the light of the rest of the galaxy. This work presents the analysis of the NSCs in a sample of 12 dwarf galaxies in the Fornax Cluster observed with MUSE. Analysis of the stellar populations and star-formation histories reveal that all the NSCs show evidence of multiple episodes of star formation, indicating that they have built up their mass further since their initial formation. The NSCs were found to have systematically lower metallicities than their host galaxies, which is consistent with a scenario for mass-assembly through mergers with infalling globular clusters, while the presence of younger stellar populations and gas emission in the core of two galaxies is indicative of in-situ star formation. We conclude that the NSCs in these dwarf galaxies likely originated as globular clusters that migrated to the core of the galaxy which have built up their mass mainly through mergers with other infalling clusters, with gas-inflow leading to in-situ star formation playing a secondary role.

astro-ph.GA↗

arXiv2020

DropEdge: Towards Deep Graph Convolutional Networks on Node Classification

\emph{Over-fitting} and \emph{over-smoothing} are two main obstacles of developing deep Graph Convolutional Networks (GCNs) for node classification. In particular, over-fitting weakens the generalization ability on small dataset, while over-smoothing impedes model training by isolating output representations from the input features with the increase in network depth. This paper proposes DropEdge, a novel and flexible technique to alleviate both issues. At its core, DropEdge randomly removes a certain number of edges from the input graph at each training epoch, acting like a data augmenter and also a message passing reducer. Furthermore, we theoretically demonstrate that DropEdge either reduces the convergence speed of over-smoothing or relieves the information loss caused by it. More importantly, our DropEdge is a general skill that can be equipped with many other backbone models (e.g. GCN, ResGCN, GraphSAGE, and JKNet) for enhanced performance. Extensive experiments on several benchmarks verify that DropEdge consistently improves the performance on a variety of both shallow and deep GCNs. The effect of DropEdge on preventing over-smoothing is empirically visualized and validated as well. Codes are released on~\url{https://github.com/DropEdge/DropEdge}.

cs.LG↗

arXiv2020

Graph Representation Learning via Graphical Mutual Information Maximization

The richness in the content of various information networks such as social networks and communication networks provides the unprecedented potential for learning high-quality expressive representations without external supervision. This paper investigates how to preserve and extract the abundant information from graph-structured data into embedding space in an unsupervised manner. To this end, we propose a novel concept, Graphical Mutual Information (GMI), to measure the correlation between input graphs and high-level hidden representations. GMI generalizes the idea of conventional mutual information computations from vector space to the graph domain where measuring mutual information from two aspects of node features and topological structure is indispensable. GMI exhibits several benefits: First, it is invariant to the isomorphic transformation of input graphs---an inevitable constraint in many existing graph representation learning algorithms; Besides, it can be efficiently estimated and maximized by current mutual information estimation methods such as MINE; Finally, our theoretical analysis confirms its correctness and rationality. With the aid of GMI, we develop an unsupervised learning model trained by maximizing GMI between the input and output of a graph neural encoder. Considerable experiments on transductive as well as inductive node classification and link prediction demonstrate that our method outperforms state-of-the-art unsupervised counterparts, and even sometimes exceeds the performance of supervised ones.

cs.LG↗

arXiv2020

Adversarial Attack on Community Detection by Hiding Individuals

It has been demonstrated that adversarial graphs, i.e., graphs with imperceptible perturbations added, can cause deep graph models to fail on node/graph classification tasks. In this paper, we extend adversarial graphs to the problem of community detection which is much more difficult. We focus on black-box attack and aim to hide targeted individuals from the detection of deep graph community detection models, which has many applications in real-world scenarios, for example, protecting personal privacy in social networks and understanding camouflage patterns in transaction networks. We propose an iterative learning framework that takes turns to update two modules: one working as the constrained graph generator and the other as the surrogate community detection model. We also find that the adversarial graphs generated by our method can be transferred to other learning based community detection models.

cs.SI↗

arXiv2020

Graph Ordering: Towards the Optimal by Learning

Graph representation learning has achieved a remarkable success in many graph-based applications, such as node classification, link prediction, and community detection. These models are usually designed to preserve the vertex information at different granularity and reduce the problems in discrete space to some machine learning tasks in continuous space. However, regardless of the fruitful progress, for some kind of graph applications, such as graph compression and edge partition, it is very hard to reduce them to some graph representation learning tasks. Moreover, these problems are closely related to reformulating a global layout for a specific graph, which is an important NP-hard combinatorial optimization problem: graph ordering. In this paper, we propose to attack the graph ordering problem behind such applications by a novel learning approach. Distinguished from greedy algorithms based on predefined heuristics, we propose a neural network model: Deep Order Network (DON) to capture the hidden locality structure from partial vertex order sets. Supervised by sampled partial order, DON has the ability to infer unseen combinations. Furthermore, to alleviate the combinatorial explosion in the training space of DON and make the efficient partial vertex order sampling , we employ a reinforcement learning model: the Policy Network, to adjust the partial order sampling probabilities during the training phase of DON automatically. To this end, the Policy Network can improve the training efficiency and guide DON to evolve towards a more effective model automatically. Comprehensive experiments on both synthetic and real data validate that DON-RL outperforms the current state-of-the-art heuristic algorithm consistently. Two case studies on graph compression and edge partitioning demonstrate the potential power of DON-RL in real applications.

cs.SI↗