Lifelong person re-identification (LReID) is an important but challenging task that suffers from catastrophic forgetting due to significant domain gaps between training steps. Existing LReID approaches typically rely on data replay and knowledge distillation to mitigate this issue. However, data replay methods compromise data privacy by storing historical exemplars, while knowledge distillation methods suffer from limited performance due to the cumulative forgetting of undistilled knowledge. To overcome these challenges, we propose a novel paradigm that models and rehearses the distribution of the old domains to enhance knowledge consolidation during the new data learning, possessing a strong anti-forgetting capacity without storing any exemplars. Specifically, we introduce an exemplar-free LReID method called Distribution Rehearsing via Adaptive Style Kernel Learning (DASK). DASK includes a Distribution Rehearser Learning mechanism that learns to transform arbitrary distribution data into the current data style at each learning step. To enhance the style transfer capacity, an Adaptive Kernel Prediction network is explored to achieve an instance-specific distribution adjustment. Additionally, we design a Distribution Rehearsing-driven LReID Training module, which rehearses old distribution based on the new data via the old AKPNet model, achieving effective knowledge accumulation. Experimental results show our DASK outperforms the existing methods by 3.6%-6.8% and 4.5%-6.5% on seen and unseen domains, respectively.

Hypersonic vehicle-(HSV) borne synthetic aperture radar (SAR) with spiral trajectory has significant advantages, such as wide angle coverage, flexible observation geometry, and rich spatial information acquisition. However, the high mobility of the platform can cause many problems for SAR imaging. To address these issues, this letter first establishes a suitable vector model for HSV SAR with spiral trajectory. We proposed an innovative method based on 2-D Lagrange interpolation to address cross–coupling and spatial variation caused by rapid altitude and velocity changes. The performance of the method is demonstrated through simulations and real data experiments. The proposed solution has significant advancements in the application of HSV SAR.

Due to the large diving trajectory, the design of the imaging algorithm remains an exceptionally difficult task for hypersonic synthetic aperture radar (SAR). In this article, an accurate range model that accounts for the impact of higher-order motion factors on imaging performance is established. Based on the model, the signal characteristics are analyzed, which indicates that the design of the imaging approach faces greatly challenges such as severe cross-couplings and significant spatial variations (SVs). In view of these issues, a time–frequency hybrid domain imaging algorithm (TFHDIA) is proposed for hypersonic vehicle (HSV)-borne SAR systems with a large diving trajectory. The cross-couplings are significantly reduced by azimuth preprocessing. The 2-D SV in the range cell migration (RCM) and the first-order SV in the Doppler parameters (DPs) are eliminated through the fractional Fourier transform (FrFT) and the extended keystone transform (EKT). Additionally, the higher-order SVs of DPs are removed by the phase filter bank, which includes range, azimuth, and cross-coupling spatially variant phase error compensation filters. The proposed algorithm exhibits strong capabilities in eliminating SVs and effectively reducing cross-coupling, making it particularly suitable for HSV SAR systems with large diving trajectories. Simulation and actual acquired data results validate the efficacy of the proposed algorithm.

Compared with conventional synthetic aperture radar (SAR) systems, hypersonic vehicle-borne (HSV) SAR with a spiral trajectory offers unique advantages, including high maneuverability, large detection range, and diverse observation views, enabling precise three-dimensional (3-D) SAR coordinate positioning. However, the high maneuverability of the spiral trajectory poses significant challenges, such as model mismatch, image defocusing, and positioning algorithm failures. To address these challenges, a geometric vector model of HSV SAR with a spiral trajectory is established, and the corresponding signal and trajectory characteristics are analyzed. Subsequently, the 3-D coordinate positioning approach based on HSV SAR with spiral trajectory is proposed. The proposed method achieves target positioning via two-dimensional (2-D) multi-view imaging, SAR target matching, and 3-D coordinate calculation. This method is relatively novel and provides superior imaging and positioning performance. Simulation, real data, and semi-real data experiments are given to validate the effectiveness of the proposed method.

Maneuver provides flexibility for highly squinted synthetic aperture radar (SAR) and also means complicated signal characteristics in the echo, especially for ground moving target imaging (GMTIm). This article analyzes the interaction of parameters between the maneuvering platform and the moving target. The analysis suggests that three key factors should be taken into account: azimuth spectrum aliasing, Doppler centroid ambiguity, and high-order errors. To deal with these challenges, a novel GMTIm methodology for maneuvering platform is presented. The proposed approach employs an advanced parameter estimation method based on the extended generalized high-order ambiguity function (EGHAF), which enables simultaneous estimation of high-order phase coefficients across all orders in low signal-to-noise ratio (SNR) conditions through 2-D extension. Due to the estimation and compensation for higher order phases, which are usually ignored in conventional methods, the proposed method is more suitable for moving target imaging with maneuvering platforms. The superiority of the proposed approach is verified by simulation and real data results.

Hypersonic glide vehicle usually flies in a spiral trajectory to avoid being detected, reconnoitered, or jammed. However, due to its complex flight characteristics, the hypersonic glide vehicle-borne (HGV) synthetic aperture radar (SAR) faces several new challenges, including the complicated geometric model, serious cross-coupling, and significant spatial variation. To address these issues, a precise range model with large maneuvering parameters is first derived on the basis of the kinematic features, indicating that the signal properties change greatly. Then, a novel nonuniform fast Fourier transform (NUFFT)-based approach performed in the 2-D frequency domain is proposed. The cross-coupling terms are decoupled according to Lagrange mean value and the spatial variations are eliminated by 2-D NUFFT operation, which has a high depth-of-focusing and is more effective for HGV SAR with spiral trajectory. The effectiveness of the proposed approach is substantiated through a series of computer simulations and semi-real data experiments.