Endowing deep models with the ability to generalize in dynamic scenarios is of vital significance for real-world deployment, given the continuous and complex changes in data distribution. Recently, evolving domain generalization (EDG) has emerged to address distribution shifts over time, aiming to capture evolving patterns for improved model generalization. However, existing EDG methods may suffer from spurious correlations by modeling only the dependence between data and targets across domains, creating a shortcut between task-irrelevant factors and the target, which hinders generalization. To this end, we design a time-aware structural causal model (SCM) that incorporates dynamic causal factors and the causal mechanism drifts, and propose Static-DYNamic Causal Representation Learning (SYNC), an approach that effectively learns time-aware causal representations. Specifically, it integrates specially designed information-theoretic objectives into a sequential VAE framework which captures evolving patterns, and produces the desired representations by preserving intra-class compactness of causal factors both across and within domains. Moreover, we theoretically show that our method can yield the optimal causal predictor for each time domain. Results on both synthetic and real-world datasets exhibit that SYNC can achieve superior temporal generalization performance.

This paper presents a rigorous evaluation of Real-to-Sim parameter estimation approaches for fabric manipulation in robotics. The study systematically assesses three state-of-the-art approaches, namely two differential pipelines and a data-driven approach. We also devise a novel physics-informed neural network approach for physics parameter estimation. These approaches are interfaced with two simulations across multiple Real-to-Sim scenarios (lifting, wind blowing, and stretching) for five different fabric types and evaluated on three unseen scenarios (folding, fling, and shaking). We found that the simulation engines and the choice of Real-to-Sim approaches significantly impact fabric manipulation performance in our evaluation scenarios. Moreover, PINN observes superior performance in quasi-static tasks but shows limitations in dynamic scenarios. Videos and source code are available at cvas-ug.github.io/real2sim-study.

The current surge in video content highlights the tasks of moment retrieval (MR) and highlight detection (HD), which involve localizing video segments of events and predicting clip-wise saliency scores based on text queries. The recent methods, while effective, may overlook two aspects: 1) Multimodal features often show weak alignment from frozen encoders, hindering thorough semantic exploration of video clips through fine-grained cross-modal interaction. 2) Due to the absence of significant distinction between adjacent video clips, it is challenging for clip-level context modeling to accurately locate query-relevant content. To mitigate these gaps and inspired by the human routine in understanding visual events, we propose a progressive framework dubbed “what and where” to initially grasp the aligned semantics of each video clip, and then proceed to scan moment-level contextual features temporally to identify events matching the query. In the ‘what’ stage, to enable explicit alignment of modal features and achieve a thorough semantic understanding, we firstly devise the Initial Semantic Projection (ISP) loss to bring closer different modal features with similar semantics. Additionally, we develop a Clip Semantic Mining module to deeply mine the relevance of these identified semantics to the specific query (at both word- and sentence-level). In the ‘where’ stage, to enhance feature distinctiveness, we design a Multi-Context Perception module that models moment-level context. It includes an Event Context (EC) branch and a Chronological Context (CC) branch, focusing on possible query-relevant event moments and temporal moments of various lengths. Finally, extensive experiments validate the state-of-the-art performance of our W2W model on three benchmark datasets without additional pre-training. Codes are available athttps://github.com/TJUMMG/W2W.

China’s new generation of polar-orbiting meteorological satellite, FengYun-3D (FY-3D), is equipped with the Medium Resolution Spectral Imager II (MERSI-II) sensor, which greatly enhances Chinese capacity for comprehensive space environment detection and meteorological remote sensing. MERSI-II has two mid-infrared (MIR) channels with the potential to apply split-window (SW) algorithms. We have developed an improved nonlinear SW (NSW) algorithm to retrieve the ground brightness temperature (GBT)$T_{\text {gb}}^{0}$without the contribution of the solar direct beam from MERSI-II MIR data, enabling the accurate retrieval of surface bidirectional reflectivity$\rho _{b}$based on the radiative transfer equation (RTE). Compared with the traditional estimation method, this article assumed a linear relationship between$\rho _{b}$of MIR channels 20 and 21, and believed that under dry-cool atmospheric conditions,$T_{\text {gb}}^{0}$is not equal. Considering different solar zenith angles (SZAs), by iterating the combination of atmospheric and surface parameters under reasonable variations, the MODerate spectral resolution atmospheric TRANsmittance (MODTRAN) model version 5.2 was used to obtain the simulation datasets for determining the NSW coefficients. In the practical retrieval, the precise algorithm coefficients are only relevant to SZA. The$R^{2}$fit by the NSW algorithm is 0.99, and the root-mean-square error (RMSE) is less than 0.58 K for different SZAs. Under different SZAs, comparing the actual value of$\rho _{b}$with the proposed method, the bias is less than$1.57\times 10^{-3}$, and the RMSE is less than$2.21\times 10^{-3}$. A detailed sensitivity analysis found that the errors in atmospheric water vapor content (WVC), CO2 concentrations, O3 concentrations, horizontal visibilities (VISs), and instrumental noise were below the order of$10^{-4}$, and their effects on the retrieval of$\rho _{b}$were negligible. This study employed two methods to validate$T_{\text {gb}}^{0}$, thereby indirectly validating$\rho _{b}$. First, the moderate resolution imaging spectrometer (MODIS) land surface emissivity (LSE) product MYD11C1 and the ERA5-Land land surface temperature (LST) product were used to validate$T_{\text {gb}}^{0}$, yielding a bias of 0.9 K and an RMSE of 1.13 K. Second, in situ measurements from the surface radiation budget network (SURFRAD) sites were used to validate$T_{\text {gb}}^{0}$, resulting in a bias of 0.74 K and an RMSE of 1.83 K. Overall, results showed that the proposed algorithm has good accuracy and can improve its applications in relevant fields.

Large-scale and high spatiotemporal resolution sea surface temperature (SST) products can provide important support for monitoring dynamic changes in the marine environment, energy development, and assimilation models. We develop an algorithm to obtain the high spatiotemporal resolution SST product by fusion of the high spatial resolution SST product from the medium resolution spectral imager (MERSI) on the Fengyun-3E (FY-3E) satellite and the high temporal SST product from the advanced geosynchronous radiation imager (AGRI) on the Fengyun-4E (FY-4E) satellite. During data preprocessing, MERSI products have large data volumes stored in chunks, and we use the multicore computer for batch re-projection to generate the matrix of temperature values and satellite observation times. The matrix has high spatial coverage and removes the anomalous data, increasing the stability of the algorithm. AGRI data have gaps due to cloud cover, and we build a pixel-by-pixel linear fitting model to extract more valid change information from multihours data, which also improve the coverage of fusion results. During the fusion process, we calculate the spectral and temporal weighting factors in real time, divide data into regular blocks for multicores fusion to get 1 km/1 h SST fusion product covering most areas of the Eastern Hemisphere. Using Argo buoy data to verify the accuracy of the results for 15 days, the root mean square error (RMSE) is$1.089~^{\circ }$C and the average deviation is$0.867~^{\circ }$C. Multicore parallel computing makes algorithm fast and efficient. The spatiotemporal resolution of results is improved significantly and the effect of missing original data of the results is reduced.

In modern processors, speculative execution has significantly improved the performance of processors, but it has also introduced speculative execution vulnerabilities. Recent defenses are based on the delayed execution to block various speculative side channels, but we show that several of the current state-of-the-art defenses fail to block some of the available speculative side channels, and the current most secure defense introduces a performance overhead of up to 24.5%. We propose SpecTerminator, the first defense framework based on instruction classes that can comprehensively and precisely block all existing speculative side channels. In SpecTerminator, a novel speculative side channel classification scheme based on the features of secret transmission is proposed, and the sensitive instructions in the speculative window are classified and identified using optimized hardware taint tracking and instruction masking techniques to accurately determine the scope of leakage. Then, according to the execution characteristics of these instructions, dedicated delayed execution strategies, such as TLB request ignoring, selective issue, and extended delay-on-miss, are designed for each type of sensitive instruction to precisely control that these instructions are delayed only in pipeline stages that are at risk of leakage. In contrast to previous defenses based on the Gem5 simulator, we have innovatively implemented defenses against Spectre attacks based on the open-source instruction set RISC-V on an FPGA-accelerated simulation platform that is more similar to real hardware. To evaluate the security of SpecTerminator, we have replicated various existing x86-based Spectre variants on RISC-V. On SPEC 2006, SpecTerminator defends against Spectre attacks based on memory hierarchy side channels with a performance overhead of 2.6% and against all existing Spectre attacks with a performance overhead of 6.0%.