Zhewen Tan, Wenhan Yu, Jianfeng Si, Tongxin Liu, Kaiqi Guan, Huiyan Jin, Jiawen Tao, Xiaokun Yuan, Xiangzheng Zhang, Duohe Ma, Tong Yang, Lin Sun. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Capturing discriminative cues with attention mechanisms is crucial for solving the high inter-class similarity problem of person re-identification (Re-ID). Self-attention (SA) learns its own contextual information within a single sample using self-affinity between elements, and some works have demonstrated its superiority in person Re-ID. However, SA weakens some subtle semantic cues and additional visual cues such as backpacks, which makes it difficult to distinguish similar-looking persons. In this paper, we propose an internal-external context interaction (IEI) attention mechanism, which aims to exploit the interaction of inter-sample latent context information and intra-sample local context information to enhance the feature representation of each element. The mechanism is able to capture subtle differences between persons and additional visual cues using inter-sample difference information and rich detail information within the element neighborhood, improving the ability to distinguish similar persons. Based on this mechanism, we propose an internal-external context interaction network (IEINet) for extracting discriminative features from multiple dimensions. In addition, to capture more discriminative information, we propose a region-diverse loss to constrain the network. Many experiments validate the effectiveness of our IEINet and demonstrate that our approach attains state-of-the-art performance on several large-scale person Re-ID datasets.

The observation of the ionosphere is of great significance for geophysics, radio propagation mechanisms, and related engineering applications. In response to the low range and Doppler resolution of the scattering function map obtained by traditional ionosonde with fewer coherent soundings, this letter proposes a super-resolution analysis method. By utilizing the coherence between the frequency components of the echo signal and using phase weighted summation, the super-resolution synthesis in the range dimension was achieved. In the frequency dimension, the Capon spectral super-resolution estimation method was employed. In this way, super-resolution is achieved both along the range and Doppler axis of the scattering function map simultaneously. It breaks the limitations of spatial and temporal sampling rates on observation resolution. Simulation and experiment results have verified the performance of this method. With a tenfold increase in resolution, it achieves the distinction of multilayer echo traces and the identifies their subtle Doppler difference. This method provides a favorable means for efficient and accurate detection of the ionosphere.

The ionosonde is one of the most widely used ionospheric detection devices. It has significant implications for the study of space physics and wireless communication technology. To solve the problem of low resolution with traditional iososonde imaging method, this letter proposes a high-resolution 2-D spatial-frequency spectral estimation imaging method. By reconstructing the echo signal matrix and constructing the guidance vector that simultaneously reflects the range and motion characteristics of the target, high-precision estimation of the range-Doppler spectrum is achieved in 2-D Capon estimation. The analysis of experimental sporadic-E (Es) sounding data of August 13, 2021 in Wuhan ($114^{\circ } 22^{\prime } $E,$30^{\circ } 30^{\prime } $N), China, demonstrates its ability of the precise measurement for single-layer echoes and the identification ability for multilayer fine structures. This method further enhances the potential of traditional ionosonde for precise observation and analysis of the ionosphere.

WIOBSS is a low-power instrument for monitoring and researching the ionosphere. Unfortunately, as operate in the High-Frequency (HF) band, its performance suffers from a lot of unwanted interferences, which can result in low signal to interference and noise ratios (SINR) and low ionogram quality. In this letter, an analysis of the range-domain correlation of radio frequency interference (RFI) in WIOBSS is presented. Based on this, a novel eigen-subspace based filtering approach for ionospheric oblique backscatter sounding is proposed. It slides the range domain sequence into the submatrix to construct the subspace and uses an improved source number estimation approach to divide the subspace according to the mode criterion. Then, the range bins with signals are projected into the eigen-subspace to suppress RFI. Experiment results show that this method is still effective under limited snapshot conditions. It can improve the SINR by about 25 dB at some typical frequencies. Furthermore, the proposed method is more robust than typical methods while the SINR fluctuates violently.

Improving the quality of the oblique sounding ionogram under the low transmit power and miniaturized antenna array structure has always been a meaningful research content. Here we try to use the adaptive beamforming technology combined with Wuhan Multi-channel Ionospheric Sounding System (WMISS) to improve the situation effectively. First, an L-shaped antenna array is employed to receive the oblique sounding signal; subsequently, the precise angle of the arrival signal is estimated by the multiple signal classification (MUSIC) algorithm. Then three typical adaptive beamforming algorithms are employed to synthesize the multi-channel signals. The experimental results show that for the ionospheric signals of oblique incidence sounding (OIS), SNR can be improved to tens of dBs, and the interference can be significantly suppressed. Especially for the swept-frequency ionogram of OIS, the trace gets clearer and more continuous, which will provide great convenience for further interpretation and ionospheric parameter inversion. In addition, by comparing the application effects of the three typical beamforming algorithms, this letter provides a reference for selecting the synthesized methods of the ionospheric signals of OIS received by the miniaturized array.

Large-scale ionospheric irregularities, which occurred on September 8, 2017, were detected by the ionosonde and the Global Navigation Satellite System (GNSS) receiver network in this letter. The Wuhan ionospheric sounding system (WISS), deployed at Puer (PUR, 22.7°N, 101.05°E), Leshan (LSH, 29.6°N, 103.7°E), and Zhangye (ZHY, 39.4°N, 100.13°E), detected these large-scale ionospheric irregularities. The 2-D maps of the rate of the GNSS total electron content (TEC) index (ROTI) further confirmed that the large-scale ionospheric irregularities can cover a broad area (from the equator to middle latitude, ~80°E to ~120°E, ~20°N to ~45°N). The ionospheric variation recorded by the WISS shows that the large-scale ionospheric irregularities were driven by the super eastward electric field, which was formed by the eastward prompt penetration electric field during a storm superimposed on the normal prereversal enhancement during the postsunset hours. Moreover, observations by the WISS show the higher the latitudes at which the ionospheric irregularities/spread F occurred, the shorter the duration time.

A novel land-based high-frequency (HF) geolocation system is proposed. The geolocating signals are transmitted from the stations with known geographic coordinates. Because of the ionospheric refraction, the signals propagate as skywave in the over-the-horizon (OTH) manner. Therefore, a wide spatial coverage can be achieved. The receiving station uses a small 2-D antenna array to estimate the angle of arrival (AOA) of each transmitter's signal. An intersecting geolocation algorithm based on the geometric relations between the transmitters, and the receiver is described, which only relies on the coordinates of the transmission stations and the azimuth angles of the signals. A novel specific multichannel system structure is also described. It has a simple structure and low hardware cost, as well as good flexibility. The preliminary experimental results prove the feasibility of this geolocation method and its engineering application significance.