5 results about "Dependence network" patented technology

This invention discloses a context-based adaptive entropy coding method for audio, comprising the following steps: First, the quantized codebook index sequence Q and quantization side information S of the audio signal are sequentially fed into a causal convolutional context model, a cross-band attention network, and a cascaded codebook dependency network for processing, to obtain temporal context feature vectors, frequency context feature vectors, and quantization-level context feature vectors for the audio signal in the time domain, respectively. Second, the temporal context feature vectors, frequency context feature vectors, and quantization-level feature vectors are input into a neural probability prediction model for processing, estimating the conditional probability distribution of each codebook index. Third, the predicted conditional probability distributions are input into an adaptive arithmetic encoder to generate the encoded bitstream, and necessary header information is added to the encoded bitstream to form a complete bitstream. This invention can further improve the compression efficiency of audio coding.

This invention provides a multimodal spatiotemporal fusion method and system for 3D target detection, relating to the field of target detection technology. In the multimodal and spatiotemporal fusion architecture, it innovatively introduces decoding branches for each modality and a current quality signal autonomously constructed from each initial detection result. This current quality signal serves as a unified adjustment mechanism, dynamically allocating feature fusion weights in both the temporal fusion and cross-modal fusion stages, breaking the limitation of traditional fusion methods that overly rely on implicit learning. Even when facing common challenges in real-world autonomous driving scenarios such as sparse LiDAR point clouds, camera obstruction by rain and fog, and glare from strong light, it can accurately detect targets. Without adding any external supervision labels, it can significantly improve the overall robustness, stability, and reliable localization capability of 3D target detection in complex scenarios.

The purpose of the present application is to provide a method and device for verifying the switching of summer and winter time, which realizes the embedding of the traditional verification logic depending on external test tools and manual intervention into the firmware runtime system of the device, and constructs a closed-loop summer time DST self-verification mechanism with the capabilities of rule analysis, event pre-play, execution monitoring, result comparison and abnormal response. The mechanism does not depend on network connection or external debugging interface, and can still complete the integrity check of the DST switching behavior independently when the device is in a completely offline state, thereby solving the problems of time confusion, data distortion and service interruption caused by the lack of self-checking capability in the prior art.

The application belongs to the field of unmanned system cluster control and autonomous cooperation, and particularly relates to an unmanned system cluster acceleration synchronization method for multi-channel network interruption. A kind of unmanned aerial vehicle cluster system is modeled as a linear system model, and a multi-channel communication topology network and a distributed elastic controller of the unmanned aerial vehicle cluster system are constructed for the model. A Lyapunov function depending on the network interruption mode is constructed, and the exponential stability and gain conditions of the synchronization error system under different network interruption modes are established to cope with the synchronization error. The synchronization rate is characterized as an optimization problem characterized by single-channel attenuation rate and network interruption duration constraints. Then, an iterative optimization algorithm is designed to solve the controller gain under different network interruption modes to achieve accelerated synchronization. Finally, the proposed method is simulated and verified, and the results show that the method can effectively solve the problem of accelerated synchronization of unmanned system cluster under multi-channel network interruption.

The present application relates to the technical field of urban rail transit disaster prevention and mitigation, in particular to a subway station multi-system cascade failure analysis and resilience improvement method, which is implemented according to the following steps; Step 1: a three-dimensional multi-layer grid model of the subway station is established, the horizontal resolution is set, the vertical communication channels of stairs / escalators between different levels are established, and the connection position, width and slope parameters are recorded; Step 2: a two-dimensional shallow water equation set containing a multi-layer coupled source term is used for fluid mechanics calculation; Step 3: a directed dependence network G=(V,E) of four systems of power, drainage, communication and channel is constructed, and the water depth threshold of each system and the dependence edge weight between systems are set; Step 4: the system node state is updated according to the water depth distribution h(x,y,z,t) using a piecewise linear state function, and the node state update equation integrates the direct flood degradation coefficient and the dependence degradation term; The method has the advantages of strong physical reality and innovative cascade mechanism, reveals the importance-vulnerability paradox and comprehensively evaluates stability.