Physical adversarial attacks (PAAs) pose a serious security threat to millimeter-wave (mmWave) radar systems used in safety-critical applications such as autonomous driving and security checking. These attacks, manipulating radar signals via specially crafted materials, are proven feasible and can cause severe sensing failures; however, effective defenses remain unexplored due to the difficulty of distinguishing adversarial examples from normal environmental objects. This paper presents mmGuard, a physics-based defense framework that addresses this challenge by exploiting a fundamental insight: the engineering process that makes materials adversarial inevitably creates detectable physical signatures. We identify three key domains where adversarial examples show artificial nature: spatial phase discontinuities, anomalous radar cross-section patterns, and violations of natural physico-kinematic relationships. mmGuard systematically captures these signatures through multi-domain feature extraction, enhances their discriminability via neural refinement, and enables efficient per-object attack detection and mitigation compatible with automotive radar update rates. To enable evaluation, we introduce mmAD, comprising over 110,000 annotated radar frames with diverse adversarial examples across realistic deployment scenarios. Experimental results demonstrate that mmGuard achieves over 90% detection accuracy while exhibiting strong in-distribution performance, with few-shot adaptation enabling calibration to unseen settings Case studies further validate that mmGuard can reliably defend against PAAs in real-world settings.

Human-centric perception with radio frequency (RF) signals has recently entered a new era of end-to-end processing with Transformers. Considering the long-sequence nature of RF signals, the State Space Model (SSM) has emerged as a superior alternative due to its effective long-sequence modeling and linear complexity. However, integrating SSM into RF-based sensing presents unique challenges including the fundamentally different signal representation, distinct frequency responses in different scenarios, and incomplete capture caused by specular reflection. To address this, we carefully devise a dual-branch SSM block that is characterized by adaptively grasping the most informative frequency cues and the assistant spatial information to fully explore the human representations from radar echoes. Based on these two branchs, we further introduce an SSM-based network for handling various downstream human perception tasks, named RFMamba. Extensive experimental results demonstrate the superior performance of our proposed RFMamba across all three downstream tasks. To the best of our knowledge, RFMamba is the first attempt to introduce SSM into RF-based human-centric perception.

Learning-based video denoisers have attained state-of-the-art (SOTA) performances on public evaluation benchmarks. Nevertheless, they typically encounter significant performance drops when applied to unseen real-world data, owing to inherent data discrepancies. To address this problem, this work delves into the model pretraining techniques and proposes masked central frame modeling (MCFM), a new video pretraining approach that significantly improves the generalization ability of the denoiser. This proposal stems from a key observation: pretraining denoiser by reconstructing intact videos from the corrupted sequences, where the central frames are masked at a suitable probability, contributes to achieving superior performance on real-world data. Building upon MCFM, we introduce a robust video denoiser, named MVDenoiser, which is firstly pretrained on massive available ordinary videos for general video modeling, and then finetuned on costful real-world noisy/clean video pairs for noisy-to-clean mapping. Additionally, beyond the denoising model, we further establish a new paired real-world noisy video dataset (RNVD) to facilitate cross-dataset evaluation of generalization ability. Extensive experiments conducted across different datasets demonstrate that the proposed method achieves superior performance compared to existing methods. Code and dataset are available athttps://github.com/mxxx99/MVDenoiser.

Recently, learning-based multi-exposure fusion (MEF) methods have made significant improvements. However, these methods mainly focus on static scenes and are prone to generate ghosting artifacts when tackling a more common scenario, i.e., the input images include motion, due to the lack of a benchmark dataset and solution for dynamic scenes. In this paper, we fill this gap by creating an MEF dataset of dynamic scenes, which contains multi-exposure image sequences and their corresponding high-quality reference images. To construct such a dataset, we propose a 'static-for-dynamic' strategy to obtain multi-exposure sequences with motions and their corresponding reference images. To the best of our knowledge, this is the first MEF dataset of dynamic scenes. Correspondingly, we propose a deep dynamic MEF (DDMEF) framework to reconstruct a ghost-free high-quality image from only two differently exposed images of a dynamic scene. DDMEF is achieved through two steps: pre-enhancement-based alignment and privilege-information-guided fusion. The former pre-enhances the input images before alignment, which helps to address the misalignments caused by the significant exposure difference. The latter introduces a privilege distillation scheme with an information attention transfer loss, which effectively improves the deghosting ability of the fusion network. Extensive qualitative and quantitative experimental results show that the proposed method outperforms state-of-the-art dynamic MEF methods. The source code and dataset are released at https://github.com/Tx000/Deep_dynamicMEF.