Time series generation is widely used in real-world applications such as simulation, data augmentation, and hypothesis testing. Recently, diffusion models have emerged as the de facto approach to time series generation, enabling diverse synthesis scenarios. However, the fixed standard-Gaussian diffusion prior may be ill-suited for time series data, which exhibit properties such as temporal order and fixed time points. In this paper, we propose TimeBridge, a framework that flexibly synthesizes time series data by using diffusion bridges to learn paths between a chosen prior and the data distribution. We then explore several prior designs tailored to time series synthesis. Our framework covers (i) data- and time-dependent priors for unconditional generation and (ii) scale-preserving priors for conditional generation. Experiments show that our framework with data-driven priors outperforms standard diffusion models on time series generation.

The inverted index is a widely used data structure to avoid the infeasible exhaustive search. It accelerates retrieval significantly by splitting the database into multiple disjoint sets and restricts distance computation to a small fraction of the database. Moreover, it even improves search quality by allowing quantizers to exploit the compact distribution of residual vector space. However, we firstly point out a problem that an existing deep learning-based quantizer hardly benefits from the residual vector space, unlike conventional shallow quantizers. To cope with this problem, we introduce a novel disentangled representation learning for unsupervised neural quantization. Similar to the concept of residual vector space, the proposed method enables more compact latent space by disentangling information of the inverted index from the vectors. Experimental results on large-scale datasets confirm that our method outperforms the state-of-the-art retrieval systems by a large margin.

Security threats increase as connectivity among vehicles increases. In particular, a lack of authentication, integrity, and confidentiality makes the controller area network (CAN) protocol, which is used in critical domains such as vehicle body and powertrain, vulnerable to threats. In this paper, we propose methods for CAN security enhancement that use a support vector machine (SVM) and the autocorrelation of the received signal to detect a malicious node. Robustness to temperature variation is also considered because autocorrelation is affected by temperature variation. There are two methods based on the degree of uniformity of the temperature distribution. If the temperature is uniformly distributed over the vehicle and the temperature sensor is embedded in the secure node, the first scheme (temperature measurement system) trains data in each segmented temperature range more precisely using multiple classifiers. If not (i.e., a nonuniform temperature distribution or an absence of a temperature sensor), the alternative scheme (all-temperature training system) trains data in all temperature ranges with a single classifier. The performances of the proposed systems are evaluated on a testbed. The proposed method can operate without modifying the CAN protocol because it is based on the characteristics of the physical layer. In addition, security can be enhanced redundantly by the system running independently without authentication protocols.