arXiv subjects

Xiang Li

Publications and source records attributed to Xiang Li.

At least 1,117 records · Page 62Linked to original sources

arXiv2022

RigNet: Repetitive Image Guided Network for Depth Completion

Depth completion deals with the problem of recovering dense depth maps from sparse ones, where color images are often used to facilitate this task. Recent approaches mainly focus on image guided learning frameworks to predict dense depth. However, blurry guidance in the image and unclear structure in the depth still impede the performance of the image guided frameworks. To tackle these problems, we explore a repetitive design in our image guided network to gradually and sufficiently recover depth values. Specifically, the repetition is embodied in both the image guidance branch and depth generation branch. In the former branch, we design a repetitive hourglass network to extract discriminative image features of complex environments, which can provide powerful contextual instruction for depth prediction. In the latter branch, we introduce a repetitive guidance module based on dynamic convolution, in which an efficient convolution factorization is proposed to simultaneously reduce its complexity and progressively model high-frequency structures. Extensive experiments show that our method achieves superior or competitive results on KITTI benchmark and NYUv2 dataset.

cs.CV↗

arXiv2022

Boosting the sensitivity of high frequency gravitational wave detectors by PT-symmetry

The kilo-Hertz gravitational waves radiated by the neutron star merger remnants carry rich information about the physics of high-density nuclear matter states, and many important astrophysical phenomena such as gamma-ray bursts and black hole formation. Current laser interferometer gravitational wave detectors, such as LIGO, VIRGO, and KAGRA have limited signal response at the kilo-Hertz band, thereby unable to capture these important physical phenomena. This work proposes an alternative protocol for boosting the sensitivity of the gravitational wave detectors at high frequency by implementing an optomechanical quantum amplifier. With the auxiliary quantum amplifier, this design has the feature of Parity-Time (PT) symmetry so that the detection band will be significantly broadened within the kilo-Hertz range. In this work, we carefully analyze the quantum-noise-limited sensitivity and the dynamical stability of this design. Based on our protocol, our result shows that the quantum-noise-limited sensitivity will be improved by one order of magnitude around 3kHz, which indicates the potential of our design for a future search of neutron star merger signals.

gr-qc↗

arXiv2022

Multi-Modal Masked Pre-Training for Monocular Panoramic Depth Completion

In this paper, we formulate a potentially valuable panoramic depth completion (PDC) task as panoramic 3D cameras often produce 360° depth with missing data in complex scenes. Its goal is to recover dense panoramic depths from raw sparse ones and panoramic RGB images. To deal with the PDC task, we train a deep network that takes both depth and image as inputs for the dense panoramic depth recovery. However, it needs to face a challenging optimization problem of the network parameters due to its non-convex objective function. To address this problem, we propose a simple yet effective approach termed M{^3}PT: multi-modal masked pre-training. Specifically, during pre-training, we simultaneously cover up patches of the panoramic RGB image and sparse depth by shared random mask, then reconstruct the sparse depth in the masked regions. To our best knowledge, it is the first time that we show the effectiveness of masked pre-training in a multi-modal vision task, instead of the single-modal task resolved by masked autoencoders (MAE). Different from MAE where fine-tuning completely discards the decoder part of pre-training, there is no architectural difference between the pre-training and fine-tuning stages in our M$^{3}$PT as they only differ in the prediction density, which potentially makes the transfer learning more convenient and effective. Extensive experiments verify the effectiveness of M{^3}PT on three panoramic datasets. Notably, we improve the state-of-the-art baselines by averagely 26.2% in RMSE, 51.7% in MRE, 49.7% in MAE, and 37.5% in RMSElog on three benchmark datasets.

cs.CV↗

arXiv2022

CGMN: A Contrastive Graph Matching Network for Self-Supervised Graph Similarity Learning

Graph similarity learning refers to calculating the similarity score between two graphs, which is required in many realistic applications, such as visual tracking, graph classification, and collaborative filtering. As most of the existing graph neural networks yield effective graph representations of a single graph, little effort has been made for jointly learning two graph representations and calculating their similarity score. In addition, existing unsupervised graph similarity learning methods are mainly clustering-based, which ignores the valuable information embodied in graph pairs. To this end, we propose a contrastive graph matching network (CGMN) for self-supervised graph similarity learning in order to calculate the similarity between any two input graph objects. Specifically, we generate two augmented views for each graph in a pair respectively. Then, we employ two strategies, namely cross-view interaction and cross-graph interaction, for effective node representation learning. The former is resorted to strengthen the consistency of node representations in two views. The latter is utilized to identify node differences between different graphs. Finally, we transform node representations into graph-level representations via pooling operations for graph similarity computation. We have evaluated CGMN on eight real-world datasets, and the experiment results show that the proposed new approach is superior to the state-of-the-art methods in graph similarity learning downstream tasks.

cs.LG↗

arXiv2022

Enhanced ultrafast X-ray diffraction by transient resonances

Diffraction-before-destruction imaging with single ultrashort X-ray pulses has the potential to visualise non-equilibrium processes, such as chemical reactions, at the nanoscale with sub-femtosecond resolution in the native environment without the need of crystallization. Here, a nanospecimen partially diffracts a single X-ray flash before sample damage occurs. The structural information of the sample can be reconstructed from the coherent X-ray interference image. State-of-art spatial resolution of such snapshots from individual heavy element nanoparticles is limited to a few nanometers. Further improvement of spatial resolution requires higher image brightness which is ultimately limited by bleaching effects of the sample. We compared snapshots from individual 100 nm Xe nanoparticles as a function of the X-ray pulse duration and incoming X-ray intensity in the vicinity of the Xe M-shell resonance. Surprisingly, images recorded with few femtosecond and sub-femtosecond pulses are up to 10 times brighter than the static linear model predicts. Our Monte-Carlo simulation and statistical analysis of the entire data set confirms these findings and attributes the effect to transient resonances. Our simulation suggests that ultrafast form factor changes during the exposure can increase the brightness of X-ray images by several orders of magnitude. Our study guides the way towards imaging with unprecedented combination of spatial and temporal resolution at the nanoscale.

physics.optics↗

arXiv2022

DTG-SSOD: Dense Teacher Guidance for Semi-Supervised Object Detection

The Mean-Teacher (MT) scheme is widely adopted in semi-supervised object detection (SSOD). In MT, the sparse pseudo labels, offered by the final predictions of the teacher (e.g., after Non Maximum Suppression (NMS) post-processing), are adopted for the dense supervision for the student via hand-crafted label assignment. However, the sparse-to-dense paradigm complicates the pipeline of SSOD, and simultaneously neglects the powerful direct, dense teacher supervision. In this paper, we attempt to directly leverage the dense guidance of teacher to supervise student training, i.e., the dense-to-dense paradigm. Specifically, we propose the Inverse NMS Clustering (INC) and Rank Matching (RM) to instantiate the dense supervision, without the widely used, conventional sparse pseudo labels. INC leads the student to group candidate boxes into clusters in NMS as the teacher does, which is implemented by learning grouping information revealed in NMS procedure of the teacher. After obtaining the same grouping scheme as the teacher via INC, the student further imitates the rank distribution of the teacher over clustered candidates through Rank Matching. With the proposed INC and RM, we integrate Dense Teacher Guidance into Semi-Supervised Object Detection (termed DTG-SSOD), successfully abandoning sparse pseudo labels and enabling more informative learning on unlabeled data. On COCO benchmark, our DTG-SSOD achieves state-of-the-art performance under various labelling ratios. For example, under 10% labelling ratio, DTG-SSOD improves the supervised baseline from 26.9 to 35.9 mAP, outperforming the previous best method Soft Teacher by 1.9 points.

cs.CV↗

arXiv2022

Online Video Instance Segmentation via Robust Context Fusion

Video instance segmentation (VIS) aims at classifying, segmenting and tracking object instances in video sequences. Recent transformer-based neural networks have demonstrated their powerful capability of modeling spatio-temporal correlations for the VIS task. Relying on video- or clip-level input, they suffer from high latency and computational cost. We propose a robust context fusion network to tackle VIS in an online fashion, which predicts instance segmentation frame-by-frame with a few preceding frames. To acquire the precise and temporal-consistent prediction for each frame efficiently, the key idea is to fuse effective and compact context from reference frames into the target frame. Considering the different effects of reference and target frames on the target prediction, we first summarize contextual features through importance-aware compression. A transformer encoder is adopted to fuse the compressed context. Then, we leverage an order-preserving instance embedding to convey the identity-aware information and correspond the identities to predicted instance masks. We demonstrate that our robust fusion network achieves the best performance among existing online VIS methods and is even better than previously published clip-level methods on the Youtube-VIS 2019 and 2021 benchmarks. In addition, visual objects often have acoustic signatures that are naturally synchronized with them in audio-bearing video recordings. By leveraging the flexibility of our context fusion network on multi-modal data, we further investigate the influence of audios on the video-dense prediction task, which has never been discussed in existing works. We build up an Audio-Visual Instance Segmentation dataset, and demonstrate that acoustic signals in the wild scenarios could benefit the VIS task.

cs.CV↗

arXiv2022

Benchmarking of DL Libraries and Models on Mobile Devices

Deploying deep learning (DL) on mobile devices has been a notable trend in recent years. To support fast inference of on-device DL, DL libraries play a critical role as algorithms and hardware do. Unfortunately, no prior work ever dives deep into the ecosystem of modern DL libs and provides quantitative results on their performance. In this paper, we first build a comprehensive benchmark that includes 6 representative DL libs and 15 diversified DL models. We then perform extensive experiments on 10 mobile devices, which help reveal a complete landscape of the current mobile DL libs ecosystem. For example, we find that the best-performing DL lib is severely fragmented across different models and hardware, and the gap between those DL libs can be rather huge. In fact, the impacts of DL libs can overwhelm the optimizations from algorithms or hardware, e.g., model quantization and GPU/DSP-based heterogeneous computing. Finally, atop the observations, we summarize practical implications to different roles in the DL lib ecosystem.

cs.LG↗

arXiv2022

RAW-GNN: RAndom Walk Aggregation based Graph Neural Network

Graph-Convolution-based methods have been successfully applied to representation learning on homophily graphs where nodes with the same label or similar attributes tend to connect with one another. Due to the homophily assumption of Graph Convolutional Networks (GCNs) that these methods use, they are not suitable for heterophily graphs where nodes with different labels or dissimilar attributes tend to be adjacent. Several methods have attempted to address this heterophily problem, but they do not change the fundamental aggregation mechanism of GCNs because they rely on summation operators to aggregate information from neighboring nodes, which is implicitly subject to the homophily assumption. Here, we introduce a novel aggregation mechanism and develop a RAndom Walk Aggregation-based Graph Neural Network (called RAW-GNN) method. The proposed approach integrates the random walk strategy with graph neural networks. The new method utilizes breadth-first random walk search to capture homophily information and depth-first search to collect heterophily information. It replaces the conventional neighborhoods with path-based neighborhoods and introduces a new path-based aggregator based on Recurrent Neural Networks. These designs make RAW-GNN suitable for both homophily and heterophily graphs. Extensive experimental results showed that the new method achieved state-of-the-art performance on a variety of homophily and heterophily graphs.

cs.LG↗

arXiv2022

Scalable Simulation of Quantum Measurement Process with Quantum Computers

Recent development in quantum information sciences and technologies, especially building programmable quantum computers, provide us new opportunities to study fundamental aspects of quantum mechanics. We propose qubit models to emulate the quantum measurement process, in which the quantum information of a qubit is mapped to a collection of qubits acting as the measurement device. One model is motivated by single-photon detection and the other by spin measurement. Both models are scalable to generate Schrödinger cat-like state, and their corresponding quantum circuits are shown explicitly. Large-scale simulations could be realized in near-term quantum computers, while classical computers cannot perform the same task efficiently. Due to the scalability of the models, such simulations can help explore the quantum-to-classical boundary, if exists, in the quantum measurement problem. Besides, our protocol to generate cat states may have important applications in quantum computing and metrology.

quant-ph↗

arXiv2022

Dual Power Spectrum Manifold and Toeplitz HPD Manifold: Enhancement and Analysis for Matrix CFAR Detection

Recently, an innovative matrix CFAR detection scheme based on information geometry, also referred to as the geometric detector, has been developed speedily and exhibits distinct advantages in several practical applications. These advantages benefit from the geometry of the Toeplitz Hermitian positive definite (HPD) manifold $\mathcal{M}_{\mathcal{T}H_{++}}$, but the sophisticated geometry also results in some challenges for geometric detectors, such as the implementation of the enhanced detector to improve the SCR (signal-to-clutter ratio) and the analysis of the detection performance. To meet these challenges, this paper develops the dual power spectrum manifold $\mathcal{M}_{\text{P}}$ as the dual space of $\mathcal{M}_{\mathcal{T}H_{++}}$. For each affine invariant geometric measure on $\mathcal{M}_{\mathcal{T}H_{++}}$, we show that there exists an equivalent function named induced potential function on $\mathcal{M}_{\text{P}}$. By the induced potential function, the measurements of the dissimilarity between two matrices can be implemented on $\mathcal{M}_{\text{P}}$, and the geometric detectors can be reformulated as the form related to the power spectrum. The induced potential function leads to two contributions: 1) The enhancement of the geometric detector, which is formulated as an optimization problem concerning $\mathcal{M}_{\mathcal{T}H_{++}}$, is transformed to an equivalent and simpler optimization on $\mathcal{M}_{\text{P}}$. In the presented example of the enhancement, the closed-form solution, instead of the gradient descent method, is provided through the equivalent optimization. 2) The detection performance is analyzed based on $\mathcal{M}_{\text{P}}$, and the advantageous characteristics, which benefit the detection performance, can be deduced by analyzing the corresponding power spectrum to the maximal point of the induced potential function.

eess.SP↗

arXiv2022

Bear the Query in Mind: Visual Grounding with Query-conditioned Convolution

Visual grounding is a task that aims to locate a target object according to a natural language expression. As a multi-modal task, feature interaction between textual and visual inputs is vital. However, previous solutions mainly handle each modality independently before fusing them together, which does not take full advantage of relevant textual information while extracting visual features. To better leverage the textual-visual relationship in visual grounding, we propose a Query-conditioned Convolution Module (QCM) that extracts query-aware visual features by incorporating query information into the generation of convolutional kernels. With our proposed QCM, the downstream fusion module receives visual features that are more discriminative and focused on the desired object described in the expression, leading to more accurate predictions. Extensive experiments on three popular visual grounding datasets demonstrate that our method achieves state-of-the-art performance. In addition, the query-aware visual features are informative enough to achieve comparable performance to the latest methods when directly used for prediction without further multi-modal fusion.

cs.CV↗

arXiv2022

Robust Inverse Framework using Knowledge-guided Self-Supervised Learning: An application to Hydrology

Machine Learning is beginning to provide state-of-the-art performance in a range of environmental applications such as streamflow prediction in a hydrologic basin. However, building accurate broad-scale models for streamflow remains challenging in practice due to the variability in the dominant hydrologic processes, which are best captured by sets of process-related basin characteristics. Existing basin characteristics suffer from noise and uncertainty, among many other things, which adversely impact model performance. To tackle the above challenges, in this paper, we propose a novel Knowledge-guided Self-Supervised Learning (KGSSL) inverse framework to extract system characteristics from driver and response data. This first-of-its-kind framework achieves robust performance even when characteristics are corrupted. We show that KGSSL achieves state-of-the-art results for streamflow modeling for CAMELS (Catchment Attributes and MEteorology for Large-sample Studies) which is a widely used hydrology benchmark dataset. Specifically, KGSSL outperforms other methods by up to 16 \% in reconstructing characteristics. Furthermore, we show that KGSSL is relatively more robust to distortion than baseline methods, and outperforms the baseline model by 35\% when plugging in KGSSL inferred characteristics.

cs.LG↗

arXiv2022

Seamless Interaction Design with Coexistence and Cooperation Modes for Robust Human-Robot Collaboration

A robot needs multiple interaction modes to robustly collaborate with a human in complicated industrial tasks. We develop a Coexistence-and-Cooperation (CoCo) human-robot collaboration system. Coexistence mode enables the robot to work with the human on different sub-tasks independently in a shared space. Cooperation mode enables the robot to follow human guidance and recover failures. A human intention tracking algorithm takes in both human and robot motion measurements as input and provides a switch on the interaction modes. We demonstrate the effectiveness of CoCo system in a use case analogous to a real world multi-step assembly task.

cs.RO↗

arXiv2022

Hybrid Instance-aware Temporal Fusion for Online Video Instance Segmentation

Recently, transformer-based image segmentation methods have achieved notable success against previous solutions. While for video domains, how to effectively model temporal context with the attention of object instances across frames remains an open problem. In this paper, we propose an online video instance segmentation framework with a novel instance-aware temporal fusion method. We first leverages the representation, i.e., a latent code in the global context (instance code) and CNN feature maps to represent instance- and pixel-level features. Based on this representation, we introduce a cropping-free temporal fusion approach to model the temporal consistency between video frames. Specifically, we encode global instance-specific information in the instance code and build up inter-frame contextual fusion with hybrid attentions between the instance codes and CNN feature maps. Inter-frame consistency between the instance codes are further enforced with order constraints. By leveraging the learned hybrid temporal consistency, we are able to directly retrieve and maintain instance identities across frames, eliminating the complicated frame-wise instance matching in prior methods. Extensive experiments have been conducted on popular VIS datasets, i.e. Youtube-VIS-19/21. Our model achieves the best performance among all online VIS methods. Notably, our model also eclipses all offline methods when using the ResNet-50 backbone.

cs.CV↗

arXiv2022

Ab initio calculation of real solids via neural network ansatz

Neural networks have been applied to tackle many-body electron correlations for small molecules and physical models in recent years. Here we propose a new architecture that extends molecular neural networks with the inclusion of periodic boundary conditions to enable ab initio calculation of real solids. The accuracy of our approach is demonstrated in four different types of systems, namely the one-dimensional periodic hydrogen chain, the two-dimensional graphene, the three-dimensional lithium hydride crystal, and the homogeneous electron gas, where the obtained results, e.g. total energies, dissociation curves, and cohesive energies, outperform many traditional ab initio methods and reach the level of the most accurate approaches. Moreover, electron densities of typical systems are also calculated to provide physical intuition of various solids. Our method of extending a molecular neural network to periodic systems can be easily integrated into other neural network structures, highlighting a promising future of ab initio solution of more complex solid systems using neural network ansatz, and more generally endorsing the application of machine learning in materials simulation and condensed matter physics.

physics.chem-ph↗

arXiv2022

Uniform Masking: Enabling MAE Pre-training for Pyramid-based Vision Transformers with Locality

Masked AutoEncoder (MAE) has recently led the trends of visual self-supervision area by an elegant asymmetric encoder-decoder design, which significantly optimizes both the pre-training efficiency and fine-tuning accuracy. Notably, the success of the asymmetric structure relies on the "global" property of Vanilla Vision Transformer (ViT), whose self-attention mechanism reasons over arbitrary subset of discrete image patches. However, it is still unclear how the advanced Pyramid-based ViTs (e.g., PVT, Swin) can be adopted in MAE pre-training as they commonly introduce operators within "local" windows, making it difficult to handle the random sequence of partial vision tokens. In this paper, we propose Uniform Masking (UM), successfully enabling MAE pre-training for Pyramid-based ViTs with locality (termed "UM-MAE" for short). Specifically, UM includes a Uniform Sampling (US) that strictly samples $1$ random patch from each $2 \times 2$ grid, and a Secondary Masking (SM) which randomly masks a portion of (usually $25\%$) the already sampled regions as learnable tokens. US preserves equivalent elements across multiple non-overlapped local windows, resulting in the smooth support for popular Pyramid-based ViTs; whilst SM is designed for better transferable visual representations since US reduces the difficulty of pixel recovery pre-task that hinders the semantic learning. We demonstrate that UM-MAE significantly improves the pre-training efficiency (e.g., it speeds up and reduces the GPU memory by $\sim 2\times$) of Pyramid-based ViTs, but maintains the competitive fine-tuning performance across downstream tasks. For example using HTC++ detector, the pre-trained Swin-Large backbone self-supervised under UM-MAE only in ImageNet-1K can even outperform the one supervised in ImageNet-22K. The codes are available at https://github.com/implus/UM-MAE.

cs.CV↗

arXiv2022

2D Toda $τ$ Functions, Weighted Hurwitz Numbers and the Cayley Graph: Determinant Representation and Recursion Formula

We generalize the determinant representation of the KP $τ$ functions to the case of the 2D Toda $τ$ functions. The generating functions for the weighted Hurwitz numbers are a parametric family of 2D Toda $τ$ functions; for which we give a determinant representation of weighted Hurwitz numbers. Then we can get a finite-dimensional equation system for the weighted Hurwitz numbers $H^d_{G}(σ,ω)$ with the same dimension $|σ|=|ω|=n$. Using this equation system, we calculated the value of the weighted Hurwitz numbers with dimension $0,\,1,\,2$ and give a recursion formula to calculating the higher dimensional weighted Hurwitz numbers. For any given weighted generating function $G(z)$, the weighted Hurwitz number degenerates into the Hurwitz numbers when $d=0$. We get a matrix representation for the Hurwitz numbers. The generating functions of weighted paths in the Cayley graph of the symmetric group are a parametric family of 2D Toda $τ$ functions; for which we obtain a determinant representation of weighted paths in the Cayley graph.

math-ph↗