9results about How to "Avoid redundant calculations" patented technology

This application provides a speech synthesis method, a vocoder, and its training and optimization method. The vocoder uses a ConvNext neural network as its backbone network. ConvNext replaces transposed convolution with channel upsampling, avoiding the periodic noise of transposed convolution. It employs depthwise convolution and separable convolution, which not only captures the time-frequency information of the vocoder input but also significantly reduces the computational load of the backbone network. Using frequency band fusion technology, the vocoder can simultaneously output audio from different frequency components of the synthesized audio, thereby reducing the output sampling rate and improving the vocoder's operating efficiency. Specifically, the vocoder uses multi-channel output, with each channel corresponding to an audio signal within a specific frequency band. The frequency ranges of the signals from each channel are orthogonal to each other, and the sampling rate is 1 / M of the original (M is the number of frequency bands). Finally, the audio signals from each frequency band are anti-mirror filtered and superimposed to obtain the audio at the original sampling rate.

This invention proposes a method and system for diagnosing open-circuit faults in three-phase inverter IGBTs, relating to the technical field of power electronic device fault diagnosis. The method involves lightweight processing of acquired historical data to obtain feature vectors; training a pre-defined three-phase inverter IGBT open-circuit fault diagnosis model based on these feature vectors; real-time acquisition of three-phase current data to be processed, followed by bounded normalization to obtain normalized current amplitudes; and generating IGBT open-circuit fault judgment results based on these normalized current amplitudes. Real-time data related to the three-phase inverter IGBT open-circuit fault is collected based on the IGBT open-circuit fault judgment results, and this real-time data is input into the trained three-phase inverter IGBT open-circuit fault diagnosis model to output the three-phase inverter IGBT open-circuit fault diagnosis results. This invention effectively achieves lightweight, efficient, and accurate diagnosis of three-phase inverter IGBT open-circuit faults.

This invention provides a rapid landslide area segmentation method based on high-resolution satellite remote sensing data and deep learning algorithms, belonging to the field of ground scene technology. The invention downsamples the original satellite remote sensing image to obtain a first image, extracts reflectance data from the red and near-infrared channels, calculates the reflectance difference and sum of reflectance values, and normalizes them to obtain a vegetation index to construct a vegetation layer. The first image and the vegetation layer are superimposed and stitched together, and feature enhancement is performed to obtain a vegetation feature map. This map is input into a convolutional neural network to output a landslide probability map. A two-dimensional coordinate set is constructed and cropped to obtain high-resolution test patches. The spatial gradient magnitude of the local vegetation index is calculated and weighted by a preset enhancement coefficient to generate edge-enhanced patches. The edge-enhanced patches are input into a semantic segmentation network to output a local landslide map, which is then backfilled into the initial image to obtain a landslide area segmentation map. This invention constructs a cascaded segmentation system to achieve efficient and accurate segmentation of landslide areas.

This application provides an intelligent monitoring method and system for engineering construction, relating to the field of engineering construction. The method includes: collecting multimodal raw data streams; converting the raw data streams into pulse events, each pulse event containing event type, timestamp, spatial coordinates, and semantic tags; binding pulse events of different modalities into spatially correlated pulse event clusters, and analyzing the temporal characteristics of these pulse event clusters in real time using a spiking neural network, triggering data processing tasks only when abnormal patterns or critical events are detected; dynamically adjusting sensor thresholds and event binding rules using the temporal dependence plasticity of pulses; binding the bound pulse event clusters to BIM model components to generate digital twins with spatiotemporal tags, and generating control commands through the digital twins. This solves the technical problems of redundant consumption caused by indiscriminate data processing and the formation of data silos by independent modules.

The invention relates to the technical field of biological information processing, in particular to a bar code correction method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring a to-be-corrected target bar code; generating a mode character string at each base position included in the target bar code, and replacing the base at the current position with the mode character string to obtain a preset number of new target bar codes; matching the new target bar code with each mode candidate set constructed by the bar code white list, and after the target mode candidate set is matched, matching the new target bar code with each new reference bar code in the target mode candidate set to obtain a first matching result; and correcting the target bar code based on the first matching result, and determining an inclusion relation between the target bar code and the bar code white list. According to the embodiment of the invention, index mapping is realized through the position mode, redundant calculation for generating a certain number of mutants is avoided, the query complexity is reduced, and the query speed is increased.

This invention specifically relates to an optical target tracking method based on trajectory guidance and spatiotemporal joint decision-making, belonging to the field of computer vision and video information processing technology. The method first extracts image features through an anchor-free twin network architecture, outputting classification, centrality scores, and regression response maps; then, it constructs a sliding window weighted least squares trajectory prediction model, modeling target motion as locally uniformly accelerated motion to achieve target position prediction; subsequently, it generates an adaptive Gaussian heatmap based on trajectory prediction uncertainty, constructing a spatial attention mechanism for trajectory guidance; finally, it fuses visual features and prior trajectory information through a linear weighted fusion strategy to output the target position. This invention effectively utilizes temporal motion information, solving the problems of low tracking accuracy and poor robustness in existing technologies under occlusion and fast-moving scenarios, balancing real-time performance with low computational overhead, and is suitable for ordinary optical and infrared thermal imaging tracking scenarios.

The application discloses a battery fault detection method and system based on multi-modal fusion and semi-supervised learning. The method comprises the following steps: obtaining battery original operation data, preprocessing the battery original operation data to obtain time series signal sample data; extracting the time domain dynamic characteristics of the time series signal sample data based on a one-dimensional convolutional neural network, and obtaining the time domain characteristic vector through global average pooling and flattening; calculating the statistical characteristics of the time series signal sample and converting them into structured text, extracting semantic features based on a pre-trained LLM to obtain a semantic feature vector; performing feature fusion based on a cross-attention mechanism to obtain a fusion feature vector, and inputting the fusion feature vector into a double-output branch to output a classification probability and a prediction confidence; and comparing the prediction confidence with a preset threshold to obtain a battery fault detection result. The application only needs normal battery management data for model training, the detection time is short, multi-modal data can be fused, and the accuracy of battery fault detection is effectively improved.

The present application is directed to a multi-feature financial domain named entity recognition method for some long-length, difficult-to-bound and various forms of expression named entities in the Chinese financial field. Firstly, the method divides entity boundary and entity classification as two independent tasks, introduces a multi-channel attention mechanism based on gate, divides the financial entity boundary on the basis of the pre-trained language model, and learns the enhanced Chinese character features. Then, the boundary division result is input into the pre-trained language model in the form of MASK for data enhancement. Subsequently, an enhanced feature based on the document-level entity is introduced to construct a financial entity classification model. Through experiments, the best-performing Chinese pre-trained language model is determined from several state-of-the-art models, which is then embedded into the method of the present application and compared with other benchmark models. The experimental results show that the proposed model is superior to other benchmark models in the named entity recognition task in the financial field.

The application relates to the technical field of farmland detection, in particular to a sky-ground integrated farmland detection system and method based on artificial intelligence, which comprises the following steps: acquiring multi-source remote sensing data, performing terrain self-adaptive preprocessing on the multi-source remote sensing data; performing terrain weighted fusion on the preprocessed multi-source remote sensing data to generate a detection layer; extracting farmland boundary features based on the detection layer; discriminating farmland shape types based on the extracted farmland boundary features and in combination with the slope grade; classifying farmland types in combination with the aspect category, the slope grade and crop growth features obtained from the multi-source remote sensing data; and dynamically adjusting unmanned aerial vehicle flight parameters and sensor arrangement density according to the slope grade, the aspect category, the farmland shape type and the farmland type. The application can realize terrain self-adaptive accurate detection of farmland in hilly and mountainous areas, and can improve detection accuracy and efficiency by deeply fusing terrain features.