Wei Jiang

In this paper, we propose a tensor-based positioning scheme for indoor WiFi systems assisted by multiple reconfigurable intelligent surfaces (RISs), leveraging deep learning (DL)-based denoising networks. This scheme aligns with the growing integration of artificial intelligence (AI)-driven innovations and advanced sensing technologies, aiming to enhance positioning accuracy and environmental mapping in smart city applications. To compensate for WiFi antenna capacity limitations, we first utilize a virtual antenna array (VAA) technique that leverages extensive carrier frequency resources to expand spatial capabilities. Considering the limited robustness of the conventional methods to noise, the proposed scheme then incorporates a tensor-based complex deep residual convolutional neural network (T-CDRCNN) for enhanced channel estimation. Next, the received signal reflected by RISs is modeled as a denoised fourth-order tensor, and factor matrices are derived through parallel factor (PARAFAC) tensor decomposition. Using the geometric relations between extracted channel parameters and the target, we ultimately estimate the user equipment (UE) location and orientation, as well as perform environmental mapping. Simulation results show that the proposed scheme demonstrates superior localization accuracy relative to existing methods, reaching centimeter-level precision, and exhibits reliable performance even at low SNRs owing to the integration of the denoising network.

The composition of soundtracks for videos is a highly integrated process that synthesizes artistic creativity with technical expertise. It involves the precise synchronization of musical elements with the visual component to effectively enhance the artistic quality of the video. Herein, we aimed to generate harmonious and temporally aligned music based on video content, assisting artists in soundtrack composition. Existing approaches predominantly generate music based on human motion data from fixed-length videos, limiting their applicability. To overcome this limitation, we propose an autoregressive unit diffusion model, $\mathbf {W^{2}}$, for variable-length video-to-music generation. $\mathbf {W^{2}}$ segments videos and music into unified temporal units and employs a diffusion model to generate music units conditioned on structured video information and contextual music information. It then autoregressively generates coherent and consistent music sequences. This model employs structured video condition information, obtained through global encoding, to effectively generate music units that consistently align with the video content. At the same time, contextual music condition information is used to ensure the generation of a continuous sequence of music units matching the duration of the video. Unlike prior models relying on human skeletal features, $\mathbf {W^{2}}$ can be applied to not only dance videos but also free-style videos. Experimental results show that $\mathbf {W^{2}}$ generates music that is temporally synchronized and semantically aligned with visual content.