arXiv subjects

Ye Yuan

Publications and source records attributed to Ye Yuan.

At least 235 records · Page 13Linked to original sources

arXiv2022

Edge-based Local Push for Personalized PageRank

Personalized PageRank (PPR) is a popular node proximity metric in graph mining and network research. Given a graph G=(V,E) and a source node $s \in V$, a single-source PPR (SSPPR) query asks for the PPR value $\vpi(u)$ with respect to s, which represents the relative importance of node u in the context of the source node s. Among existing algorithms for SSPPR queries, LocalPush is a fundamental method which serves as a cornerstone for subsequent algorithms. In LocalPush, a push operation is a crucial primitive operation, which distributes the probability at a node u to ALL u's neighbors via the corresponding edges. Although this push operation works well on unweighted graphs, unfortunately, it can be rather inefficient on weighted graphs. In particular, on unbalanced weighted graphs where only a few of these edges take the majority of the total weight among them, the push operation would have to distribute insignificant probabilities along those edges which just take the minor weights, resulting in expensive overhead. To resolve this issue, we propose the EdgePush algorithm, a novel method for computing SSPPR queries on weighted graphs. EdgePush decomposes the aforementioned push operations in edge-based push, allowing the algorithm to operate at the edge level granularity. Hence, it can flexibly distribute the probabilities according to edge weights. Furthermore, our EdgePush allows a fine-grained termination threshold for each individual edge, leading to a superior complexity over LocalPush. Notably, we prove that EdgePush improves the theoretical query cost of LocalPush by an order of up to O(n) when the graph's weights are unbalanced, both in terms of $\ell_1$-error and normalized additive error. Our experimental results demonstrate that EdgePush significantly outperforms state-of-the-art baselines in terms of query efficiency on large motif-based and real-world weighted graphs.

cs.DS↗

arXiv2022

A Sampling Theorem for Exact Identification of Continuous-time Nonlinear Dynamical Systems

Low sampling frequency challenges the exact identification of the continuous-time (CT) dynamical system from sampled data, even when its model is identifiable. The necessary and sufficient condition is proposed -- which is built from Koopman operator -- to the exact identification of the CT system from sampled data. The condition gives a Nyquist-Shannon-like critical frequency for exact identification of CT nonlinear dynamical systems with Koopman invariant subspaces: 1) it establishes a sufficient condition for a sampling frequency that permits a discretized sequence of samples to discover the underlying system and 2) it also establishes a necessary condition for a sampling frequency that leads to system aliasing that the underlying system is indistinguishable; and 3) the original CT signal does not have to be band-limited as required in the Nyquist-Shannon Theorem. The theoretical criterion has been demonstrated on a number of simulated examples, including linear systems, nonlinear systems with equilibria, and limit cycles.

eess.SY↗

arXiv2022

Self-Supervised Information Bottleneck for Deep Multi-View Subspace Clustering

In this paper, we explore the problem of deep multi-view subspace clustering framework from an information-theoretic point of view. We extend the traditional information bottleneck principle to learn common information among different views in a self-supervised manner, and accordingly establish a new framework called Self-supervised Information Bottleneck based Multi-view Subspace Clustering (SIB-MSC). Inheriting the advantages from information bottleneck, SIB-MSC can learn a latent space for each view to capture common information among the latent representations of different views by removing superfluous information from the view itself while retaining sufficient information for the latent representations of other views. Actually, the latent representation of each view provides a kind of self-supervised signal for training the latent representations of other views. Moreover, SIB-MSC attempts to learn the other latent space for each view to capture the view-specific information by introducing mutual information based regularization terms, so as to further improve the performance of multi-view subspace clustering. To the best of our knowledge, this is the first work to explore information bottleneck for multi-view subspace clustering. Extensive experiments on real-world multi-view data demonstrate that our method achieves superior performance over the related state-of-the-art methods.

cs.LG↗

arXiv2022

Unified Simulation, Perception, and Generation of Human Behavior

Understanding and modeling human behavior is fundamental to almost any computer vision and robotics applications that involve humans. In this thesis, we take a holistic approach to human behavior modeling and tackle its three essential aspects -- simulation, perception, and generation. Throughout the thesis, we show how the three aspects are deeply connected and how utilizing and improving one aspect can greatly benefit the other aspects. We also discuss the lessons learned and our vision for what is next for human behavior modeling.

cs.CV↗

arXiv2022

Transform2Act: Learning a Transform-and-Control Policy for Efficient Agent Design

An agent's functionality is largely determined by its design, i.e., skeletal structure and joint attributes (e.g., length, size, strength). However, finding the optimal agent design for a given function is extremely challenging since the problem is inherently combinatorial and the design space is prohibitively large. Additionally, it can be costly to evaluate each candidate design which requires solving for its optimal controller. To tackle these problems, our key idea is to incorporate the design procedure of an agent into its decision-making process. Specifically, we learn a conditional policy that, in an episode, first applies a sequence of transform actions to modify an agent's skeletal structure and joint attributes, and then applies control actions under the new design. To handle a variable number of joints across designs, we use a graph-based policy where each graph node represents a joint and uses message passing with its neighbors to output joint-specific actions. Using policy gradient methods, our approach enables joint optimization of agent design and control as well as experience sharing across different designs, which improves sample efficiency substantially. Experiments show that our approach, Transform2Act, outperforms prior methods significantly in terms of convergence speed and final performance. Notably, Transform2Act can automatically discover plausible designs similar to giraffes, squids, and spiders. Code and videos are available at https://sites.google.com/view/transform2act.

cs.LG↗

arXiv2022

Online No-regret Model-Based Meta RL for Personalized Navigation

The interaction between a vehicle navigation system and the driver of the vehicle can be formulated as a model-based reinforcement learning problem, where the navigation systems (agent) must quickly adapt to the characteristics of the driver (environmental dynamics) to provide the best sequence of turn-by-turn driving instructions. Most modern day navigation systems (e.g, Google maps, Waze, Garmin) are not designed to personalize their low-level interactions for individual users across a wide range of driving styles (e.g., vehicle type, reaction time, level of expertise). Towards the development of personalized navigation systems that adapt to a variety of driving styles, we propose an online no-regret model-based RL method that quickly conforms to the dynamics of the current user. As the user interacts with it, the navigation system quickly builds a user-specific model, from which navigation commands are optimized using model predictive control. By personalizing the policy in this way, our method is able to give well-timed driving instructions that match the user's dynamics. Our theoretical analysis shows that our method is a no-regret algorithm and we provide the convergence rate in the agnostic setting. Our empirical analysis with 60+ hours of real-world user data using a driving simulator shows that our method can reduce the number of collisions by more than 60%.

cs.LG↗

arXiv2022

A Piecewise Learning Framework for Control of Unknown Nonlinear Systems with Stability Guarantees

We propose a piecewise learning framework for controlling nonlinear systems with unknown dynamics. While model-based reinforcement learning techniques in terms of some basis functions are well known in the literature, when it comes to more complex dynamics, only a local approximation of the model can be obtained using a limited number of bases. The complexity of the identifier and the controller can be considerably high if obtaining an approximation over a larger domain is desired. To overcome this limitation, we propose a general piecewise nonlinear framework where each piece is responsible for locally learning and controlling over some region of the domain. We obtain rigorous uncertainty bounds for the learned piecewise models. The piecewise affine (PWA) model is then studied as a special case, for which we propose an optimization-based verification technique for stability analysis of the closed-loop system. Accordingly, given a time-discretization of the learned {PWA} system, we iteratively search for a common piecewise Lyapunov function in a set of positive definite functions, where a non-monotonic convergence is allowed. This Lyapunov candidate is verified on the uncertain system to either provide a certificate for stability or find a counter-example when it fails. This counter-example is added to a set of samples to facilitate the further learning of a Lyapunov function. We demonstrate the results on two examples and show that the proposed approach yields a less conservative region of attraction (ROA) compared with alternative state-of-the-art approaches. Moreover, we provide the runtime results to demonstrate potentials of the proposed framework in real-world implementations.

math.OC↗

arXiv2022

GLAMR: Global Occlusion-Aware Human Mesh Recovery with Dynamic Cameras

We present an approach for 3D global human mesh recovery from monocular videos recorded with dynamic cameras. Our approach is robust to severe and long-term occlusions and tracks human bodies even when they go outside the camera's field of view. To achieve this, we first propose a deep generative motion infiller, which autoregressively infills the body motions of occluded humans based on visible motions. Additionally, in contrast to prior work, our approach reconstructs human meshes in consistent global coordinates even with dynamic cameras. Since the joint reconstruction of human motions and camera poses is underconstrained, we propose a global trajectory predictor that generates global human trajectories based on local body movements. Using the predicted trajectories as anchors, we present a global optimization framework that refines the predicted trajectories and optimizes the camera poses to match the video evidence such as 2D keypoints. Experiments on challenging indoor and in-the-wild datasets with dynamic cameras demonstrate that the proposed approach outperforms prior methods significantly in terms of motion infilling and global mesh recovery.

cs.CV↗

arXiv2022

Snowmass2021 Whitepaper: Muonium to antimuonium conversion

The spontaneous muonium to antimuonium conversion is one of the interesting charged lepton flavor violation processes. It serves as a clear indication of new physics and plays an important role in constraining the parameter space beyond Standard Model. MACE is a proposed experiment to probe such a phenomenon and expected to enhance the sensitivity to the conversion probability by more than two orders of magnitude from the current best upper constraint obtained by the PSI experiment two decades ago. Recent developments in the theoretical and experimental aspects to search for such a rare process are summarized.

hep-ph↗

arXiv2022

Disentangling kinetics from thermodynamics in heterogeneous colloidal systems

Nucleation and growth (N&G) - the emergence of a new phase within an initially homogeneous one - is one of the most important physical phenomena by which gas-liquid, liquid-liquid and solid-liquid phase separation takes place. Accordingly, thermodynamics sets the asymptotic boundaries towards which the system must evolve, while kinetics tries to cope with it by imposing the transport rates at which phase separation is realized. In all heterogeneous colloidal systems observed in nature, the composition, shape, structure and ultimately physical properties result from the trade-off between thermodynamics and kinetics. In this work we demonstrate, by carefully selecting colloidal systems and controlling phase separation in microfluidic devices, that it becomes possible to go beyond N&G, disentangling kinetics effects from thermodynamics in composition, structure and physical properties of the final system. Using amyloid fibril and cellulose nanocrystal filamentous colloids for which the binodal curve defining the two-phase region in the phase diagram is given by two separate vertical lines, we extrude a solution set at one thermodynamic branch inside the other branch, realizing nematic or cholesteric droplets where the composition is set by thermodynamics, while the structure and morphology are defined by dynamic flow parameters. We demonstrate that departing from the N&G paradigm unveils new physical phenomena, such as orders of magnitude shorter timescales, a wider phase diagram and internal cholesteric structures that are not observable via conventional LLPS. We also show that by co-dispersing plasmonic gold nanoparticles within colloidal liquid crystalline droplets, our approach enables on-demand fabrication of multicomponent heterogeneous liquid crystals, enhancing their potential, and introducing original fundamental and technological directions in multicomponent structured fluids.

cond-mat.soft↗

arXiv2021

Boosting Contrastive Learning with Relation Knowledge Distillation

While self-supervised representation learning (SSL) has proved to be effective in the large model, there is still a huge gap between the SSL and supervised method in the lightweight model when following the same solution. We delve into this problem and find that the lightweight model is prone to collapse in semantic space when simply performing instance-wise contrast. To address this issue, we propose a relation-wise contrastive paradigm with Relation Knowledge Distillation (ReKD). We introduce a heterogeneous teacher to explicitly mine the semantic information and transferring a novel relation knowledge to the student (lightweight model). The theoretical analysis supports our main concern about instance-wise contrast and verify the effectiveness of our relation-wise contrastive learning. Extensive experimental results also demonstrate that our method achieves significant improvements on multiple lightweight models. Particularly, the linear evaluation on AlexNet obviously improves the current state-of-art from 44.7% to 50.1%, which is the first work to get close to the supervised 50.5%. Code will be made available.

cs.CV↗

arXiv2021

Drive high power UVC-LED wafer into low-cost 4-inch era: effect of strain modulation

Ultraviolet-C light-emitting diodes (UVC-LEDs) have great application in pathogen inactivation under various kinds of situations, especially in the fight against the COVID-19. Unfortunately, its epitaxial wafers are so far limited to 2-inch size, which greatly increases the cost of massive production. In this work, we report the 4-inch crack-free high-power UVC-LED wafer. This achievement relies on a proposed strain-tailored strategy, where a three-dimensional to two-dimensional (3D-2D) transition layer is introduced during the homo-epitaxy of AlN on high temperature annealed (HTA)-AlN template, which successfully drives the original compressive strain into tensile one and thus solves the challenge of realizing high quality Al$_{0.6}$Ga$_{0.4}$N layer with a flat surface. This smooth Al$_{0.6}$Ga$_{0.4}$N layer is nearly pseudomorphically grown on the strain-tailored HTA-AlN template, leading to 4-inch UVC-LED wafers with outstanding performances. Our strategy succeeds in compromising the bottlenecked contradictory in producing large-sized UVC-LED wafer on pronounced crystalline AlN template: The compressive strain in HTA-AlN allows for crack-free 4-inch wafer, but at the same time leads to a deterioration of the AlGaN morphology and crystal quality. The launch of 4-inch wafers makes the chip fabrication process of UVC-LEDs matches the mature blue one, and will definitely speed up the universal of UVC-LED in daily life.

cond-mat.mtrl-sci↗

arXiv2021

Approximate Graph Propagation

Efficient computation of node proximity queries such as transition probabilities, Personalized PageRank, and Katz are of fundamental importance in various graph mining and learning tasks. In particular, several recent works leverage fast node proximity computation to improve the scalability of Graph Neural Networks (GNN). However, prior studies on proximity computation and GNN feature propagation are on a case-by-case basis, with each paper focusing on a particular proximity measure. In this paper, we propose Approximate Graph Propagation (AGP), a unified randomized algorithm that computes various proximity queries and GNN feature propagation, including transition probabilities, Personalized PageRank, heat kernel PageRank, Katz, SGC, GDC, and APPNP. Our algorithm provides a theoretical bounded error guarantee and runs in almost optimal time complexity. We conduct an extensive experimental study to demonstrate AGP's effectiveness in two concrete applications: local clustering with heat kernel PageRank and node classification with GNNs. Most notably, we present an empirical study on a billion-edge graph Papers100M, the largest publicly available GNN dataset so far. The results show that AGP can significantly improve various existing GNN models' scalability without sacrificing prediction accuracy.

cs.DS↗

arXiv2021

Deep Unsupervised Active Learning on Learnable Graphs

Recently deep learning has been successfully applied to unsupervised active learning. However, current method attempts to learn a nonlinear transformation via an auto-encoder while ignoring the sample relation, leaving huge room to design more effective representation learning mechanisms for unsupervised active learning. In this paper, we propose a novel deep unsupervised Active Learning model via Learnable Graphs, named ALLG. ALLG benefits from learning optimal graph structures to acquire better sample representation and select representative samples. To make the learnt graph structure more stable and effective, we take into account $k$-nearest neighbor graph as a priori, and learn a relation propagation graph structure. We also incorporate shortcut connections among different layers, which can alleviate the well-known over-smoothing problem to some extent. To the best of our knowledge, this is the first attempt to leverage graph structure learning for unsupervised active learning. Extensive experiments performed on six datasets demonstrate the efficacy of our method.

cs.LG↗

arXiv2021

FBNet: Feature Balance Network for Urban-Scene Segmentation

Image segmentation in the urban scene has recently attracted much attention due to its success in autonomous driving systems. However, the poor performance of concerned foreground targets, e.g., traffic lights and poles, still limits its further practical applications. In urban scenes, foreground targets are always concealed in their surrounding stuff because of the special camera position and 3D perspective projection. What's worse, it exacerbates the unbalance between foreground and background classes in high-level features due to the continuous expansion of the reception field. We call it Feature Camouflage. In this paper, we present a novel add-on module, named Feature Balance Network (FBNet), to eliminate the feature camouflage in urban-scene segmentation. FBNet consists of two key components, i.e., Block-wise BCE(BwBCE) and Dual Feature Modulator(DFM). BwBCE serves as an auxiliary loss to ensure uniform gradients for foreground classes and their surroundings during backpropagation. At the same time, DFM intends to enhance the deep representation of foreground classes in high-level features adaptively under the supervision of BwBCE. These two modules facilitate each other as a whole to ease feature camouflage effectively. Our proposed method achieves a new state-of-the-art segmentation performance on two challenging urban-scene benchmarks, i.e., Cityscapes and BDD100K. Code will be released for reproduction.

cs.CV↗

arXiv2021

DeceFL: A Principled Decentralized Federated Learning Framework

Traditional machine learning relies on a centralized data pipeline, i.e., data are provided to a central server for model training. In many applications, however, data are inherently fragmented. Such a decentralized nature of these databases presents the biggest challenge for collaboration: sending all decentralized datasets to a central server raises serious privacy concerns. Although there has been a joint effort in tackling such a critical issue by proposing privacy-preserving machine learning frameworks, such as federated learning, most state-of-the-art frameworks are built still in a centralized way, in which a central client is needed for collecting and distributing model information (instead of data itself) from every other client, leading to high communication pressure and high vulnerability when there exists a failure at or attack on the central client. Here we propose a principled decentralized federated learning algorithm (DeceFL), which does not require a central client and relies only on local information transmission between clients and their neighbors, representing a fully decentralized learning framework. It has been further proven that every client reaches the global minimum with zero performance gap and achieves the same convergence rate $O(1/T)$ (where $T$ is the number of iterations in gradient descent) as centralized federated learning when the loss function is smooth and strongly convex. Finally, the proposed algorithm has been applied to a number of applications to illustrate its effectiveness for both convex and nonconvex loss functions, demonstrating its applicability to a wide range of real-world medical and industrial applications.

cs.LG↗

arXiv2021

Scaling Up Distance-generalized Core Decomposition

Core decomposition is a fundamental operator in network analysis. In this paper, we study the problem of computing distance-generalized core decomposition on a network. A distance-generalized core, also termed $(k, h)$-core, is a maximal subgraph in which every vertex has at least $k$ other vertices at distance no larger than $h$. The state-of-the-art algorithm for solving this problem is based on a peeling technique which iteratively removes the vertex (denoted by $v$) from the graph that has the smallest $h$-degree. The $h$-degree of a vertex $v$ denotes the number of other vertices that are reachable from $v$ within $h$ hops. Such a peeling algorithm, however, needs to frequently recompute the $h$-degrees of $v$'s neighbors after deleting $v$, which is typically very costly for a large $h$. To overcome this limitation, we propose an efficient peeling algorithm based on a novel $h$-degree updating technique. Instead of recomputing the $h$-degrees, our algorithm can dynamically maintain the $h$-degrees for all vertices via exploring a very small subgraph, after peeling a vertex. We show that such an $h$-degree updating procedure can be efficiently implemented by an elegant bitmap technique. In addition, we also propose a sampling-based algorithm and a parallelization technique to further improve the efficiency. Finally, we conduct extensive experiments on 12 real-world graphs to evaluate our algorithms. The results show that, when $h\ge 3$, our exact and sampling-based algorithms can achieve up to $10\times$ and $100\times$ speedup over the state-of-the-art algorithm, respectively.

cs.DS↗

arXiv2021

AgentFormer: Agent-Aware Transformers for Socio-Temporal Multi-Agent Forecasting

Predicting accurate future trajectories of multiple agents is essential for autonomous systems, but is challenging due to the complex agent interaction and the uncertainty in each agent's future behavior. Forecasting multi-agent trajectories requires modeling two key dimensions: (1) time dimension, where we model the influence of past agent states over future states; (2) social dimension, where we model how the state of each agent affects others. Most prior methods model these two dimensions separately, e.g., first using a temporal model to summarize features over time for each agent independently and then modeling the interaction of the summarized features with a social model. This approach is suboptimal since independent feature encoding over either the time or social dimension can result in a loss of information. Instead, we would prefer a method that allows an agent's state at one time to directly affect another agent's state at a future time. To this end, we propose a new Transformer, AgentFormer, that jointly models the time and social dimensions. The model leverages a sequence representation of multi-agent trajectories by flattening trajectory features across time and agents. Since standard attention operations disregard the agent identity of each element in the sequence, AgentFormer uses a novel agent-aware attention mechanism that preserves agent identities by attending to elements of the same agent differently than elements of other agents. Based on AgentFormer, we propose a stochastic multi-agent trajectory prediction model that can attend to features of any agent at any previous timestep when inferring an agent's future position. The latent intent of all agents is also jointly modeled, allowing the stochasticity in one agent's behavior to affect other agents. Our method substantially improves the state of the art on well-established pedestrian and autonomous driving datasets.

cs.AI↗