arXiv subjects
Publications and source records attributed to Jia Shen.
In this note, we study the random data problem for incompressible Navier-Stokes equations in Euclidean space with arbitrary regularity, using the randomization introduced in \cite{Shen-Soffer-Wu-2021-1}.
This paper is based on the background of the 2nd Wireless Communication Artificial Intelligence (AI) Competition (WAIC) which is hosted by IMT-2020(5G) Promotion Group 5G+AIWork Group, where the framework of the eigenvector-based channel state information (CSI) feedback problem is firstly provided. Then a basic Transformer backbone for CSI feedback referred to EVCsiNet-T is proposed. Moreover, a series of potential enhancements for deep learning based (DL-based) CSI feedback including i) data augmentation, ii) loss function design, iii) training strategy, and iv) model ensemble are introduced. The experimental results involving the comparison between EVCsiNet-T and traditional codebook methods over different channels are further provided, which show the advanced performance and a promising prospect of Transformer on DL-based CSI feedback problem.
In this work, we consider the 3D defocusing energy-critical nonlinear Schrödinger equation $i\partial_t u+Δu =|u|^4 u,\quad (t,x)\in \mathbb{R}\times \mathbb{R}^3$. Applying the outgoing and incoming decomposition presented in the recent work \cite{BECEANU-DENG-SOFFER-WU-2021}, we prove that any radial function $f$ with $χ_{\leq1}f\in H^1$ and $χ_{\geq1}f\in H^{s_0}$ with $\frac{5}{6}<s_0<1$, there exists an outgoing component $f_+$ (or incoming component $f_-$) of $f$, such that when the initial data is $f_+$, then the corresponding solution is globally well-posed and scatters forward in time; when the initial data is $f_-$, then the corresponding solution is globally well-posed and scatters backward in time.
In this paper, we study the defocusing energy-critical nonlinear Schrödinger equations $$ i\partial_t u + Δu = |u|^{\frac{4}{d-2}} u. $$ When $d=3,4$, we prove the almost sure scattering for the equations with non-radial data in $H_x^s$ for any $s\in\mathbb{R}$. In particular, our result does not rely on any spherical symmetry, size or regularity restrictions.
In this paper, we give a systematic description of the 1st Wireless Communication Artificial Intelligence (AI) Competition (WAIC) which is hosted by IMT-2020(5G) Promotion Group 5G+AI Work Group. Firstly, the framework of full channel state information (F-CSI) feedback problem and its corresponding channel dataset are provided. Then the enhancing schemes for DL-based F-CSI feedback including i) channel data analysis and preprocessing, ii) neural network design and iii) quantization enhancement are elaborated. The final competition results composed of different enhancing schemes are presented. Based on the valuable experience of 1st WAIC, we also list some challenges and potential study areas for the design of AI-based wireless communication systems.
In this paper, we consider the following three dimensional defocusing cubic nonlinear Schrödinger equation (NLS) with partial harmonic potential \begin{equation*}\tag{NLS} i\partial_t u + \left(Δ_{\mathbb{R}^3 }-x^2 \right) u = |u|^2 u, \quad u|_{t=0} = u_0. \end{equation*} Our main result shows that the solution $u$ scatters for any given initial data $u_0$ with finite mass and energy. The main new ingredient in our approach is to approximate (NLS) in the large-scale case by a relevant dispersive continuous resonant (DCR) system. The proof of global well-posedness and scattering of the new (DCR) system is greatly inspired by the fundamental works of Dodson \cite{D3,D1,D2} in his study of scattering for the mass-critical nonlinear Schrödinger equation. The analysis of (DCR) system allows us to utilize the additional regularity of the smooth nonlinear profile so that the celebrated concentration-compactness/rigidity argument of Kenig and Merle applies.
Shadow detection in a single image has received significant research interest in recent years. However, much fewer works have been explored in shadow detection over dynamic scenes. The bottleneck is the lack of a well-established dataset with high-quality annotations for video shadow detection. In this work, we collect a new video shadow detection dataset, which contains 120 videos with 11, 685 frames, covering 60 object categories, varying lengths, and different motion/lighting conditions. All the frames are annotated with a high-quality pixel-level shadow mask. To the best of our knowledge, this is the first learning-oriented dataset for video shadow detection. Furthermore, we develop a new baseline model, named triple-cooperative video shadow detection network (TVSD-Net). It utilizes triple parallel networks in a cooperative manner to learn discriminative representations at intra-video and inter-video levels. Within the network, a dual gated co-attention module is proposed to constrain features from neighboring frames in the same video, while an auxiliary similarity loss is introduced to mine semantic information between different videos. Finally, we conduct a comprehensive study on ViSha, evaluating 12 state-of-the-art models (including single image shadow detectors, video object segmentation, and saliency detection methods). Experiments demonstrate that our model outperforms SOTA competitors.
Real-world applications require light-weight, energy-efficient, fully autonomous robots. Yet, increasing autonomy is oftentimes synonymous with escalating computational requirements. It might thus be desirable to offload intensive computation--not only sensing and planning, but also low-level whole-body control--to remote servers in order to reduce on-board computational needs. Fifth Generation (5G) wireless cellular technology, with its low latency and high bandwidth capabilities, has the potential to unlock cloud-based high performance control of complex robots. However, state-of-the-art control algorithms for legged robots can only tolerate very low control delays, which even ultra-low latency 5G edge computing can sometimes fail to achieve. In this work, we investigate the problem of cloud-based whole-body control of legged robots over a 5G link. We propose a novel approach that consists of a standard optimization-based controller on the network edge and a local linear, approximately optimal controller that significantly reduces on-board computational needs while increasing robustness to delay and possible loss of communication. Simulation experiments on humanoid balancing and walking tasks that includes a realistic 5G communication model demonstrate significant improvement of the reliability of robot locomotion under jitter and delays likely to experienced in 5G wireless links.
We revisit the scattering problems for the 2D mass super-critical Schrödinger and Klein-Gordon equations with radial data below the ground state in the energy space. We give an alternative proof of energy scattering for both defocusing and focusing cases using the ideas of Dodson-Murphy \citep{dodson2017new-radial}. Our results also include the exponential type nonlinearities which seems to be new for the focusing exponential NLS.
We study the scattering problems for the quadratic Klein-Gordon equations with radial initial data in the energy space. For 3D, we prove small data scattering, and for 4D, we prove large data scattering with mass below the ground state.