Membership testing, which determines whether an element belongs to a set, is widely used in fields like database systems and network applications. Bloom Filters (BFs) can solve this problem efficiently but suffer from high False Positive Rates (FPRs) and large memory requirements for massive datasets. Learned Bloom Filters (LBFs), combining a learning model with a backup Bloom Filter, mitigate these issues by capturing data distributions. However, the critical problem of memory allocation between the learning model and the backup filter has usually been overlooked, despite its significant impact on LBF performance under constrained budgets.

Cloud Block Storage (CBS) systems underpin modern cloud infrastructures by decoupling storage from computation and enabling resource pooling for elasticity and cost efficiency. However, CBS faces two persistent challenges: load imbalance across storage nodes and network & storage amplification caused by redundancy mechanisms. While recent hybrid-redundancy block storage (HRBS) architectures combine replication caches with EC layers to reduce amplification, their static cache admission policies fail to adapt to dynamic cluster conditions, especially during node failures, leading to severe load imbalance and degraded throughput.

In IoT (Internet of Things) scenarios, massive floating-point time series data are generated in a streaming manner and transmitted within limited bandwidth for real-time analysis. To enhance the efficiency, it is acknowledged to compress the data before transmission. Existing floating-point compression methods are either for batched compression that may cause long delays, or for streaming lossless compression that has an unsatisfactory compression ratio when certain errors are allowed. In this paper, we propose the first Streaming ERror-bounded Floating-point compression Serf , which has two implementations: Serf-Qt and Serf-XOR . Serf-Qt first quantizes each floating-point value into an integer, and then encodes the integer with Elias gamma coding. Serf-XOR is the first lossy floating-point compression based on the XORing operation. To enhance the compression ratio of Serf-XOR , we propose a novel data offset technique to increase the leading zeros of the XORed values, and design a novel approximation technique to search for an error-qualified value that produces an XORed value with many trailing zeros. To improve the compression efficiency, we propose a pruning strategy to accelerate the process of approximated values search. We further build a streaming transmission prototype system based on a real development board, and deploy the proposed methods to it. Extensive experiments using 13 datasets show that, compared with 17 competitors, both Serf-Qt and Serf-XOR enjoy remarkable compression ratios with high efficiency in streaming scenarios. The transmission experiments based on the proposed system also showcase that Serf-XOR always takes the least overall time when the bandwidth is limited.

Multispectral object detection, which combines RGB visible light and thermal infrared spectral information, has broad applications in complex environments and varying illumination conditions. However, existing methods face challenges in processing multispectral data, such as inconspicuous object features in spectral images and significant discrepancies between input modality spaces and output detection spaces. To address these issues, we propose an innovative multispectral object detection method that combines contrastive learning and a new cross-modal feature fusion module. We introduce a mask feature contrastive loss that maximizes the similarity between the box-level mask features and modal features while suppressing background responses, enabling effective representative alignment between the input and output spaces. Additionally, we propose a mask-guided attention fusion module that uses a predicted pseudo mask to guide the fusion of different modal features, enhancing object responses and reducing background noise interference. Our extensive experiments on several challenging multispectral datasets demonstrate that our proposed COFNet achieves state-of-the-art performance.