Searcharxiv⌕ Search

arXiv subjects

Liang Song

Publications and source records attributed to Liang Song.

At least 55 records · Page 3Linked to original sources

Local smoothing and Hardy spaces for Fourier integral operators on manifolds

We introduce the Hardy spaces for Fourier integral operators on Riemannian manifolds with bounded geometry. We then use these spaces to obtain improved local smoothing estimates for Fourier integral operators satisfying the cinematic curvature condition, and for wave equations on compact manifolds. The estimates are essentially sharp, for all $2<p<\infty$ and on each compact manifold. We also apply our local smoothing estimates to nonlinear wave equations with initial data outside of $L^{2}$-based Sobolev spaces.

math.AP↗

Spatio-Temporal Domain Awareness for Multi-Agent Collaborative Perception

Multi-agent collaborative perception as a potential application for vehicle-to-everything communication could significantly improve the perception performance of autonomous vehicles over single-agent perception. However, several challenges remain in achieving pragmatic information sharing in this emerging research. In this paper, we propose SCOPE, a novel collaborative perception framework that aggregates the spatio-temporal awareness characteristics across on-road agents in an end-to-end manner. Specifically, SCOPE has three distinct strengths: i) it considers effective semantic cues of the temporal context to enhance current representations of the target agent; ii) it aggregates perceptually critical spatial information from heterogeneous agents and overcomes localization errors via multi-scale feature interactions; iii) it integrates multi-source representations of the target agent based on their complementary contributions by an adaptive fusion paradigm. To thoroughly evaluate SCOPE, we consider both real-world and simulated scenarios of collaborative 3D object detection tasks on three datasets. Extensive experiments demonstrate the superiority of our approach and the necessity of the proposed components.

cs.CV↗

SC-NeRF: Self-Correcting Neural Radiance Field with Sparse Views

In recent studies, the generalization of neural radiance fields for novel view synthesis task has been widely explored. However, existing methods are limited to objects and indoor scenes. In this work, we extend the generalization task to outdoor scenes, trained only on object-level datasets. This approach presents two challenges. Firstly, the significant distributional shift between training and testing scenes leads to black artifacts in rendering results. Secondly, viewpoint changes in outdoor scenes cause ghosting or missing regions in rendered images. To address these challenges, we propose a geometric correction module and an appearance correction module based on multi-head attention mechanisms. We normalize rendered depth and combine it with light direction as query in the attention mechanism. Our network effectively corrects varying scene structures and geometric features in outdoor scenes, generalizing well from object-level to unseen outdoor scenes. Additionally, we use appearance correction module to correct appearance features, preventing rendering artifacts like blank borders and ghosting due to viewpoint changes. By combining these modules, our approach successfully tackles the challenges of outdoor scene generalization, producing high-quality rendering results. When evaluated on four datasets (Blender, DTU, LLFF, Spaces), our network outperforms previous methods. Notably, compared to MVSNeRF, our network improves average PSNR from 19.369 to 25.989, SSIM from 0.838 to 0.889, and reduces LPIPS from 0.265 to 0.224 on Spaces outdoor scenes.

cs.CV↗

Learning Causality-inspired Representation Consistency for Video Anomaly Detection

Video anomaly detection is an essential yet challenging task in the multimedia community, with promising applications in smart cities and secure communities. Existing methods attempt to learn abstract representations of regular events with statistical dependence to model the endogenous normality, which discriminates anomalies by measuring the deviations to the learned distribution. However, conventional representation learning is only a crude description of video normality and lacks an exploration of its underlying causality. The learned statistical dependence is unreliable for diverse regular events in the real world and may cause high false alarms due to overgeneralization. Inspired by causal representation learning, we think that there exists a causal variable capable of adequately representing the general patterns of regular events in which anomalies will present significant variations. Therefore, we design a causality-inspired representation consistency (CRC) framework to implicitly learn the unobservable causal variables of normality directly from available normal videos and detect abnormal events with the learned representation consistency. Extensive experiments show that the causality-inspired normality is robust to regular events with label-independent shifts, and the proposed CRC framework can quickly and accurately detect various complicated anomalies from real-world surveillance videos.

cs.MM↗

Multi-scale Spatial-temporal Interaction Network for Video Anomaly Detection

Video Anomaly Detection (VAD) is an essential yet challenging task in signal processing. Since certain anomalies cannot be detected by isolated analysis of either temporal or spatial information, the interaction between these two types of data is considered crucial for VAD. However, current dual-stream architectures either confine this integral interaction to the bottleneck of the autoencoder or introduce anomaly-irrelevant background pixels into the interactive process, hindering the accuracy of VAD. To address these deficiencies, we propose a Multi-scale Spatial-Temporal Interaction Network (MSTI-Net) for VAD. First, to prioritize the detection of moving objects in the scene and harmonize the substantial semantic discrepancies between the two types of data, we propose an Attention-based Spatial-Temporal Fusion Module (ASTFM) as a substitute for the conventional direct fusion. Furthermore, we inject multi-ASTFM-based connections that bridge the appearance and motion streams of the dual-stream network, thus fostering multi-scale spatial-temporal interaction. Finally, to bolster the delineation between normal and abnormal activities, our system records the regular information in a memory module. Experimental results on three benchmark datasets validate the effectiveness of our approach, which achieves AUCs of 96.8%, 87.6%, and 73.9% on the UCSD Ped2, CUHK Avenue, and ShanghaiTech datasets, respectively.

cs.CV↗

Lp bounds for Stein's spherical maximal operators

Let ${\frak M}^α$ be the spherical maximal operators of complex order $α$ on ${\mathbb R^n}$. In this article we show that when $n\geq 2$, suppose \begin{eqnarray*} \|{\frak M}^α f \|_{L^p({\mathbb R^n})} \leq C\|f \|_{L^p({\mathbb R^n})} \end{eqnarray*} holds for some $α$ and $p\geq 2$, then we must have ${\rm Re}\,α\geq \max \{1/p-(n-1)/2,\ -(n-1)/p \}.$ When $n=2$, we prove that $\|{\frak M}^α f \|_{L^p({\mathbb R^2})} \leq C\|f \|_{L^p({\mathbb R^2})}$ if ${\rm Re}\ \ α>\max\{1/p-1/2,\ -1/p\}$, and hence the range of $α$ is sharp in the sense the estimate fails for ${\rm Re}\ α<\max\{1/p-1/2, -1/ p\}.$

math.AP↗

The Garnett-Jones Theorem on BMO spaces associated with operators and applications

Let $X$ be a metric space with doubling measure, and $L$ be a nonnegative self-adjoint operator on $L^2(X)$ whose heat kernel satisfies the Gaussian upper bound. Let $f$ be in the space $ {\rm BMO}_L(X)$ associated with the operator $L$ and we define its distance from the subspace $L^{\infty}(X)$ under the $ {\rm BMO}_L(X)$ norm as follows: $$ {\rm dist} (f, L^{\infty}):= \inf_{g\in L^{\infty}} \|f -g\|_{{\rm BMO}_L(X)}. $$ In this paper we prove that ${\rm dist} (f, L^{\infty})$ is equivalent to the infimum of the constant $\varepsilon$ in the John-Nirenberg inequality for the space ${\rm BMO}_L(X)$: $$ \sup_B { μ\big(\{ x\in B: |f(x)-e^{-{r_B^2}L}f(x)|>λ\}\big) \over μ(B)} \leq e^{-λ/\varepsilon}\ \ \ \ {\rm for\ large\ } λ. $$ This extends the well-known result of Garnett and Jones \cite{GJ1} for the classical ${\rm BMO}$ space (introduced by John and Nirenberg). As an application, we show that a ${\rm BMO}_L(X)$ function with compact support can be decomposed as the summation of an $L^\infty$-function and the integral of the heat kernel (associated with $L$) against a finite Carleson measure on $X\times[0,\infty)$. The key new technique is a geometric construction involving the semigroup $e^{-tL}$. We also resort to several fundamental tools including the stopping time argument and the random dyadic lattice.

math.CA↗

A novel efficient Multi-view traffic-related object detection framework

With the rapid development of intelligent transportation system applications, a tremendous amount of multi-view video data has emerged to enhance vehicle perception. However, performing video analytics efficiently by exploiting the spatial-temporal redundancy from video data remains challenging. Accordingly, we propose a novel traffic-related framework named CEVAS to achieve efficient object detection using multi-view video data. Briefly, a fine-grained input filtering policy is introduced to produce a reasonable region of interest from the captured images. Also, we design a sharing object manager to manage the information of objects with spatial redundancy and share their results with other vehicles. We further derive a content-aware model selection policy to select detection methods adaptively. Experimental results show that our framework significantly reduces response latency while achieving the same detection accuracy as the state-of-the-art methods.

cs.CV↗

Pointwise Decay of solutions to the energy critical nonlinear Schrödinger equations

In this note, we prove pointwise decay in time of solutions to the 3D energy-critical nonlinear Schrödinger equations assuming data in $L^1\cap H^3$. The main ingredients are the boundness of the Schrödinger propagators in Hardy space due to Miyachi \cite{Miyachi} and a fractional Leibniz rule in the Hardy space. We also extend the fractional chain rule to the Hardy space.

math.AP↗

Dirichlet problem for Schrödinger operators on Heisenberg groups

We investigate the Dirichlet problem associated to the Schrödinger operator $\mathcal L=-Δ_{\mathbb{H}^n}+V$ on Heisenberg group $\mathbb H^n$: \begin{align*} \begin{cases} \partial_{ss}u(g,s)-\mathcal L u(g,s)=0\,,\quad &{\rm in \,\ } \mathbb{H}^n\times\mathbb{R}^+,\\ u(g,0)=f \,,\quad &{\rm on \,\ } \mathbb{H}^n \end{cases} \end{align*} with $f$ in $L^p(\mathbb{H}^n)$ ($1< p<\infty$) and in $H^1_{\mathcal L}(\mathbb{H}^n)$, i.e., the Hardy space associated with $\mathcal L$. Here $Δ_{\mathbb{H}^n}$ is the sub-Laplacian on $\mathbb H^n$ and the nonnegative potential $V$ belongs to the reverse Hölder class $B_{Q/2}$ with $Q$ the homogeneous dimension of $\mathbb{H}^n$. The new approach is to establish a suitable weak maximum principle, which is the key to solve this problem under the condition $V\in B_{Q/2}$. This result is new even back to $\mathbb R^n$ (the condition will become $V\in B_{n/2}$) since the previous known result requires $V\in B_{(n+1)/2}$ which went through a Liouville type theorem.

math.AP↗

Characterizations of the Hardy space $\mathcal{H}_{FIO}^{1}(\mathbb{R}^{n})$ for Fourier integral operators

The Hardy spaces for Fourier integral operators $\mathcal{H}_{FIO}^{p}(\mathbb{R}^{n})$, for $1\leq p\leq \infty$, were introduced by Smith in [Smith,1998] and Hassell et al. in [Hassell-Portal-Rozendaal,2020]. In this article, we give several equivalent characterizations of $\mathcal{H}_{FIO}^{1}(\mathbb{R}^{n})$, for example in terms of Littlewood--Paley $g$ functions and maximal functions. This answers a question from [Rozendaal,2021]. We also give several applications of the characterizations.

math.AP↗

Learning Appearance-motion Normality for Video Anomaly Detection

Video anomaly detection is a challenging task in the computer vision community. Most single task-based methods do not consider the independence of unique spatial and temporal patterns, while two-stream structures lack the exploration of the correlations. In this paper, we propose spatial-temporal memories augmented two-stream auto-encoder framework, which learns the appearance normality and motion normality independently and explores the correlations via adversarial learning. Specifically, we first design two proxy tasks to train the two-stream structure to extract appearance and motion features in isolation. Then, the prototypical features are recorded in the corresponding spatial and temporal memory pools. Finally, the encoding-decoding network performs adversarial learning with the discriminator to explore the correlations between spatial and temporal patterns. Experimental results show that our framework outperforms the state-of-the-art methods, achieving AUCs of 98.1% and 89.8% on UCSD Ped2 and CUHK Avenue datasets.

cs.CV↗

Rethinking Saliency Map: An Context-aware Perturbation Method to Explain EEG-based Deep Learning Model

Deep learning is widely used to decode the electroencephalogram (EEG) signal. However, there are few attempts to specifically investigate how to explain the EEG-based deep learning models. We conduct a review to summarize the existing works explaining the EEG-based deep learning model. Unfortunately, we find that there is no appropriate method to explain them. Based on the characteristic of EEG data, we suggest a context-aware perturbation method to generate a saliency map from the perspective of the raw EEG signal. Moreover, we also justify that the context information can be used to suppress the artifacts in the EEG-based deep learning model. In practice, some users might want a simple version of the explanation, which only indicates a few features as salient points. To this end, we propose an optional area limitation strategy to restrict the highlighted region. To validate our idea and make a comparison with the other methods, we select three representative EEG-based models to implement experiments on the emotional EEG dataset DEAP. The results of the experiments support the advantages of our method.

cs.LG↗

Learning from Attacks: Attacking Variational Autoencoder for Improving Image Classification

Adversarial attacks are often considered as threats to the robustness of Deep Neural Networks (DNNs). Various defending techniques have been developed to mitigate the potential negative impact of adversarial attacks against task predictions. This work analyzes adversarial attacks from a different perspective. Namely, adversarial examples contain implicit information that is useful to the predictions i.e., image classification, and treat the adversarial attacks against DNNs for data self-expression as extracted abstract representations that are capable of facilitating specific learning tasks. We propose an algorithmic framework that leverages the advantages of the DNNs for data self-expression and task-specific predictions, to improve image classification. The framework jointly learns a DNN for attacking Variational Autoencoder (VAE) networks and a DNN for classification, coined as Attacking VAE for Improve Classification (AVIC). The experiment results show that AVIC can achieve higher accuracy on standard datasets compared to the training with clean examples and the traditional adversarial training.

cs.LG↗

The CMO-Dirichlet problem for the Schrödinger equation in the upper half-space and characterizations of CMO

Let $\mathcal{L}$ be a Schrödinger operator of the form $\mathcal{L}=-Δ+V$ acting on $L^2(\mathbb R^n)$ where the nonnegative potential $V$ belongs to the reverse Hölder class ${RH}_q$ for some $q\geq (n+1)/2$. Let ${CMO}_{\mathcal{L}}(\mathbb{R}^n)$ denote the function space of vanishing mean oscillation associated to $\mathcal{L}$. In this article we will show that a function $f$ of ${ CMO}_{\mathcal{L}}(\mathbb{R}^n) $ is the trace of the solution to $\mathbb{L}u=-u_{tt}+\mathcal{L} u=0$, $u(x,0)=f(x)$, if and only if, $u$ satisfies a Carleson condition $$ \sup_{B: \ { balls}}\mathcal{C}_{u,B} :=\sup_{B(x_B,r_B): \ { balls}} r_B^{-n}\int_0^{r_B}\int_{B(x_B, r_B)} \big|t \nabla u(x,t)\big|^2\, \frac{ dx\, dt } {t} <\infty, $$ and $$ \lim _{a \rightarrow 0}\sup _{B: r_{B} \leq a} \,\mathcal{C}_{u,B} = \lim _{a \rightarrow \infty}\sup _{B: r_{B} \geq a} \,\mathcal{C}_{u,B} = \lim _{a \rightarrow \infty}\sup _{B: B \subseteq \left(B(0, a)\right)^c} \,\mathcal{C}_{u,B}=0. $$ This continues the lines of the previous characterizations by Duong, Yan and Zhang \cite{DYZ} and Jiang and Li \cite{JL} for the ${ BMO}_{\mathcal{L}}$ spaces, which were founded by Fabes, Johnson and Neri \cite{FJN} for the classical BMO space. For this purpose, we will prove two new characterizations of the ${ CMO}_{\mathcal{L}}(\mathbb{R}^n)$ space, in terms of mean oscillation and the theory of tent spaces, respectively.

math.CA↗

BiOpt: Bi-Level Optimization for Few-Shot Segmentation

Few-shot segmentation is a challenging task that aims to segment objects of new classes given scarce support images. In the inductive setting, existing prototype-based methods focus on extracting prototypes from the support images; however, they fail to utilize semantic information of the query images. In this paper, we propose Bi-level Optimization (BiOpt), which succeeds to compute class prototypes from the query images under inductive setting. The learning procedure of BiOpt is decomposed into two nested loops: inner and outer loop. On each task, the inner loop aims to learn optimized prototypes from the query images. An init step is conducted to fully exploit knowledge from both support and query features, so as to give reasonable initialized prototypes into the inner loop. The outer loop aims to learn a discriminative embedding space across different tasks. Extensive experiments on two benchmarks verify the superiority of our proposed BiOpt algorithm. In particular, we consistently achieve the state-of-the-art performance on 5-shot PASCAL-$5^i$ and 1-shot COCO-$20^i$.

cs.CV↗

Prototype Rectification for Few-Shot Learning

Few-shot learning requires to recognize novel classes with scarce labeled data. Prototypical network is useful in existing researches, however, training on narrow-size distribution of scarce data usually tends to get biased prototypes. In this paper, we figure out two key influencing factors of the process: the intra-class bias and the cross-class bias. We then propose a simple yet effective approach for prototype rectification in transductive setting. The approach utilizes label propagation to diminish the intra-class bias and feature shifting to diminish the cross-class bias. We also conduct theoretical analysis to derive its rationality as well as the lower bound of the performance. Effectiveness is shown on three few-shot benchmarks. Notably, our approach achieves state-of-the-art performance on both miniImageNet (70.31% on 1-shot and 81.89% on 5-shot) and tieredImageNet (78.74% on 1-shot and 86.92% on 5-shot).

cs.CV↗

$L^p(\mathbb{R}^2)$-boundedness of Hilbert Transforms and Maximal Functions along Plane Curves with Two-variable Coefficients

In this paper, for general plane curves $γ$ satisfying some suitable smoothness and curvature conditions, we obtain the single annulus $L^p(\mathbb{R}^2)$-boundedness of the Hilbert transforms $H^\infty_{U,γ}$ along the variable plane curves $(t,U(x_1, x_2)γ(t))$ and the $L^p(\mathbb{R}^2)$-boundedness of the corresponding maximal functions $M^\infty_{U,γ}$, where $p>2$ and $U$ is a measurable function. The range on $p$ is sharp. Furthermore, for $1<p\leq 2$, under the additional conditions that $U$ is Lipschitz and making a $\varepsilon_0$-truncation with $γ(2 \varepsilon_0)\leq 1/4\|U\|_{\textrm{Lip}}$, we also obtain similar boundedness for these two operators $H^{\varepsilon_0}_{U,γ}$ and $M^{\varepsilon_0}_{U,γ}$.

math.CA↗