Multivariate time series anomaly detection is crucial for the reliability and stability of operations in cloud-edge computing systems. However, fast model training requirements, unlabeled datasets, and high-dimensional time series made it challenging to build a model that can quickly and accurately localize anomalies with good robustness. Therefore, this paper proposes a robust anomaly detection approach with spatiotemporal representation learning for multivariate time series (RAD-SRL) to address the above challenges. RAD-SRL leverages improved Dilated Causal Convolutions (DCNs) to extract spatiotemporal representations in MTS and incorporates a multi-head self-attention mechanism and an Update Gate Recurrent Neural Network (UGRNN) to capture multiple-dimensional information. Moreover, the self-adjusting mechanism of RAD-SRL reduces computational resource consumption and improves the generalizability of the model. Experimental studies on four public datasets demonstrate that the RAD-SRL outperforms the baseline methods in terms of both detection performance and time efficiency.

Recent latent diffusion models (LDMs) have been explored to generate diverse domain-specific images based on source domain data, showing promising performance in domain generalization tasks. However, although the generated images present counterfactual augmentation, such as the background and style changes, the distortion of object details disrupts the causal factors, such as texture and shape. This leads to negative outcomes when directly applying LDM to domain generalization in object detection. To address the problems mentioned above, we propose Object-Preserving Counterfactual Diffusion augmentation method (OPCD) to explore the diffusion model to generate diverse domain-specific images without disrupting the object details. First, we construct a region-aware image generation framework, which leverages labeled source domain data to guide LDM in generating region-constrained images that preserve the semantic consistency of the original source images. Second, we propose object-preserving counterfactual augmentation, which retains the object region of the generated image and fuses diversified global information. This ensures that object details are not distorted and that the generated information is maintained. Third, to reduce the resource burden of generating a large number of images in LDM, we design a random insertion strategy. It mixes generated and source domain images, turning limited diversity samples into abundant training data. Experimental results on several benchmark datasets show that OPCD outperforms existing methods in single-domain generalized object detection. Codes can be found at https://github.com/qinhongda8/OPCD.