Text-to-Audio-Video (T2AV) generation aims to produce temporally and semantically aligned visual and auditory content from natural language descriptions. While recent progress in text-to-audio and text-to-video models has improved generation quality within each modality, jointly modeling them remains challenging due to incomplete and asymmetric correspondence: audio often reflects only a subset of the visual scene, and vice versa. Naively enforcing full alignment introduces semantic noise and temporal mismatches. To address this, we propose a novel framework that performs selective cross-modal alignment through a learnable masking mechanism, enabling the model to isolate and align only the shared latent components relevant to both modalities. This mechanism is integrated into an adaptation module that interfaces with pretrained encoders and decoders from latent video and audio diffusion models, preserving their generative capacity with reduced training overhead. Theoretically, we show that our masked objective provably recovers the minimal set of shared latent variables across modalities. Empirically, our method achieves state-of-the-art performance on standard T2AV benchmarks, demonstrating significant improvements in audiovisual synchronization and semantic consistency.

Wireless sensor network (WSN) technology provides support infrastructure to solve blind zone coverage. Previous studies mainly focus on deployment and management of sensor nodes for enhancing signal coverage on a specific zone. But few of them are on coordinating several nodes to recognize blind zone. Moreover, it is a great challenge how to extract, transmit and process information energy efficiently in a decentralized way. In this paper, we propose a distributed algorithm for convenient and efficient blind zone recognition based on the early work on the unbiased broadcast gossip algorithm (UBGA) and the compressed sensing. Using UBGA, a faster consensus algorithm, the system can simultaneously compute the compressed data based on compressed sensing technique and disseminate them throughout the network. From these compressed data one can reconstruct the original data with approximate accuracy by querying any sensor values using the gradient projection strategy. Simulation results illustrate that our solution, proposed in this paper, can effectively recognize blind zone.