arXiv subjects

Jinshuo Zhang

Publications and source records attributed to Jinshuo Zhang.

10 recordsLinked to original sources

arXiv2026

Uni-Encoder Meets Multi-Encoders: Representation Before Fusion for Brain Tumor Segmentation with Missing Modalities

Multimodal MRI offers complementary information for brain tumor segmentation, but clinical scans often lack one or more modalities, which degrades segmentation performance. In this paper, we propose UniME (Uni-Encoder Meets Multi-Encoders), a two-stage heterogeneous method for brain tumor segmentation with missing modalities that reconciles the trade-offs among fine-grained structure capture, cross-modal complementarity modeling, and exploitation of available modalities. The idea is to decouple representation learning from segmentation via a two-stage heterogeneous architecture. Stage 1 pretrains a single ViT Uni-Encoder with masked image modeling to establish a unified representation robust to missing modalities. Stage 2 adds modality-specific CNN Multi-Encoders to extract high-resolution, multi-scale, fine-grained features. We fuse these features with the global representation to produce precise segmentations. Experiments on BraTS 2023 and BraTS 2024 show that UniME outperforms previous methods under incomplete multi-modal scenarios. The code is available at https://github.com/Hooorace-S/UniME

cs.CV↗

arXiv2026

Deep Learning-Based Multi-Satellite Massive MIMO Transmission: Centralized or Decentralized?

This paper investigates new efficient transmission architectures for multi-satellite massive multiple-input multiple-output (MIMO). We study the weighted sum-rate maximization problem in a multi-satellite system where multiple satellites transmit independent data streams to multi-antenna user terminals, thereby achieving higher throughput. We first adopt a multi-satellite weighted minimum mean square error (WMMSE) formulation under statistical channel state information (CSI), which yields closed-form updates for the precoding and receive vectors. To overcome the high complexity of optimization, we propose a learning-based WMMSE design that integrates tensor equivariance with closed-form recovery, enabling inference with near-optimal performance without iterative updates. Moreover, to reduce inter-satellite signaling overhead incurred by exchanging CSI and precoding vectors in centralized coordination, we develop a decentralized multi-satellite transmission scheme in which each satellite locally infers its precoders rather than receiving from the central satellite. The proposed decentralized scheme leverages periodically available satellite state information, such as orbital positions and satellite attitude, which is inherently accessible in satellite networks, and employs a dual-branch tensor-equivariant network to predict the precoders at each satellite locally. Numerical results demonstrate that the proposed multi-satellite transmission significantly outperforms single-satellite systems in sum rate; the decentralized scheme achieves sum-rate performance close to the centralized schemes while substantially reducing computational complexity and inter-satellite overhead; and the learning-based schemes exhibit strong robustness and scalability across different scenarios.

eess.SP↗

arXiv2025

Unlocking Symbol-Level Precoding Efficiency Through Tensor Equivariant Neural Network

Although symbol-level precoding (SLP) based on constructive interference (CI) exploitation offers performance gains, its high complexity remains a bottleneck. This paper addresses this challenge with an end-to-end deep learning (DL) framework with low inference complexity that leverages the structure of the optimal SLP solution in the closed-form and its inherent tensor equivariance (TE), where TE denotes that a permutation of the input induces the corresponding permutation of the output. Building upon the computationally efficient model-based formulations, as well as their known closed-form solutions, we analyze their relationship with linear precoding (LP) and investigate the corresponding optimality condition. We then construct a mapping from the problem formulation to the solution and prove its TE, based on which the designed networks reveal a specific parameter-sharing pattern that delivers low computational complexity and strong generalization. Leveraging these, we propose the backbone of the framework with an attention-based TE module, achieving linear computational complexity. Furthermore, we demonstrate that such a framework is also applicable to imperfect CSI scenarios, where we design a TE-based network to map the CSI, statistics, and symbols to auxiliary variables. Simulation results show that the proposed framework captures substantial performance gains of optimal SLP, while achieving an approximately 80-times speedup over conventional methods and maintaining strong generalization across user numbers and symbol block lengths.

eess.SP↗

arXiv2025

Technosignature Searches of Interstellar Objects

With the discovery of the third confirmed interstellar object (ISO), 3I/ATLAS, we have entered a new phase in the exploration of these long-predicted objects. Though confirmed discovery of ISOs is quite recent, their utility as targets in the search for technosignatures (historically known as the Search for Extraterrestrial Intelligence -- SETI) has been discussed for many decades. With the upcoming NSF-DOE Vera C. Rubin Observatory's Legacy Survey of Space and Time (LSST), the discovery and tracking of such objects is expected to become routine, and thus so must our examination of these objects for possible technosignatures. Here we review the literature surrounding ISOs as targets for technosignatures, which provides a well-developed motivation for such exploration. We outline four broad classes of technosignatures that are well suited for ISO follow-up, including the type of data needed and the best timing for study. Given the limitations in the current understanding of ISOs, we show that care must be taken in identifying technosignatures based primarily on comparison to objects in the Solar System. We therefore provide a roadmap for careful and consistent study of the population of ISOs in the hope of identifying technosignatures.

astro-ph.IM↗

arXiv2021

DAVOS: Semi-Supervised Video Object Segmentation via Adversarial Domain Adaptation

Domain shift has always been one of the primary issues in video object segmentation (VOS), for which models suffer from degeneration when tested on unfamiliar datasets. Recently, many online methods have emerged to narrow the performance gap between training data (source domain) and test data (target domain) by fine-tuning on annotations of test data which are usually in shortage. In this paper, we propose a novel method to tackle domain shift by first introducing adversarial domain adaptation to the VOS task, with supervised training on the source domain and unsupervised training on the target domain. By fusing appearance and motion features with a convolution layer, and by adding supervision onto the motion branch, our model achieves state-of-the-art performance on DAVIS2016 with 82.6% mean IoU score after supervised training. Meanwhile, our adversarial domain adaptation strategy significantly raises the performance of the trained model when applied on FBMS59 and Youtube-Object, without exploiting extra annotations.

cs.CV↗

arXiv2021

EDNet: Efficient Disparity Estimation with Cost Volume Combination and Attention-based Spatial Residual

Existing state-of-the-art disparity estimation works mostly leverage the 4D concatenation volume and construct a very deep 3D convolution neural network (CNN) for disparity regression, which is inefficient due to the high memory consumption and slow inference speed. In this paper, we propose a network named EDNet for efficient disparity estimation. Firstly, we construct a combined volume which incorporates contextual information from the squeezed concatenation volume and feature similarity measurement from the correlation volume. The combined volume can be next aggregated by 2D convolutions which are faster and require less memory than 3D convolutions. Secondly, we propose an attention-based spatial residual module to generate attention-aware residual features. The attention mechanism is applied to provide intuitive spatial evidence about inaccurate regions with the help of error maps at multiple scales and thus improve the residual learning efficiency. Extensive experiments on the Scene Flow and KITTI datasets show that EDNet outperforms the previous 3D CNN based works and achieves state-of-the-art performance with significantly faster speed and less memory consumption.

cs.CV↗

arXiv2016

Domain configurations in Co/Pd and L10-FePt nanowire arrays with perpendicular magnetic anisotropy

Perpendicular magnetic anisotropy [Co/Pd]15 and L10-FePt nanowire arrays of period 63 nm with linewidths 38 nm and 27 nm and film thickness 27 nm and 20 nm respectively were fabricated using a self-assembled PS-b-PDMS diblock copolymer film as a lithographic mask. The wires are predicted to support Neel walls in the Co/Pd and Bloch walls in the FePt. Magnetostatic interactions from nearest neighbor nanowires promote a ground state configuration consisting of alternating up and down magnetization in adjacent wires. This was observed over ~75% of the Co/Pd wires after ac-demagnetization but was less prevalent in the FePt because the ratio of interaction field to switching field was much smaller. Interactions also led to correlations in the domain wall positions in adjacent Co/Pd nanowires. The reversal process was characterized by nucleation of reverse domains, followed at higher fields by propagation of the domains along the nanowires. These narrow wires provide model system for exploring domain wall structure and dynamics in perpendicular anisotropy systems.

cond-mat.mtrl-sci↗

arXiv2016

Oersted field and spin current effects on magnetic domains in [Co/Pd]15 nanowires

An out-of-plane Oersted field produced from a current-carrying Au wire is used to induce local domain formation in wires made from [Co/Pd]15 multilayers with perpendicular anisotropy. A 100 ns pulsed current of 56-110 mA injected into the Au wire created a reverse domain size of 120-290 nm in a Co/Pd nanowire on one side of the Au wire. A Biot-Savart model was used to estimate the position dependence of the Oersted field around the Au wire. The shape, size and location of the reversed region of Co/Pd was consistent with the magnitude of the Oersted field and the switching field distribution of the unpatterned film. A current density of 6.2 x 10^11 A m-2 in the Co/Pd nanowire did not translate the domain walls due to low spin transfer efficiency, but the Joule heating promoted domain growth in a field below the coercive field.

cond-mat.mtrl-sci↗

arXiv2015

Edge-modulated perpendicular magnetic anisotropy in [Co/Pd]n and L10-FePt thin film wires

Thickness modulation at the edges of nanostructured magnetic thin films is shown to have important effects on their perpendicular magnetic anisotropy. Thin film wires with tapered edges were made from [Co/Pd]20 multilayers or L10-FePt films using liftoff with a double layer resist. The effect of edge taper on the reversal process was studied using magnetic force microscopy and micromagnetic modeling. In [Co/Pd]20 the anisotropy was lower in the tapered edge regions which switched at a lower reverse field compared to the center of the wire. The L10-FePt wires showed opposite behavior with the tapered regions exhibiting higher anisotropy.

cond-mat.mes-hall↗

arXiv2015

360° Domain Walls: Stability, Magnetic Field and Electric Current Effects

The formation of 360° magnetic domain walls (360DWs) in Co and Ni80Fe20 thin film wires was demonstrated experimentally for different wire widths, by successively injecting two 180° domain walls (180DWs) into the wire. For narrow wires (less than 50 nm wide for Co), edge roughness prevented the combination of the 180DWs into a 360DW, and for wide wires (200 nm for Co) the 360DW collapsed, but over an intermediate range of wire widths, reproducible 360DW formation occurred. The annihilation and dissociation of 360DWs was demonstrated by applying a magnetic field parallel to the wire, showing that annihilation fields were several times higher than dissociation fields in agreement with micromagnetic modeling. The annihilation of a 360DW by current pulsing was demonstrated.

cond-mat.mes-hall↗