8results about How to "Eliminate individual differences" patented technology

The invention provides a posture adjustment method and device, electronic equipment, a storage medium and a program product, and relates to the technical field of animation.The posture adjustment method comprises the steps that position data of a first target skeleton and a second target skeleton of a virtual character are obtained from skeleton data of the virtual character, the first target skeleton and the second target skeleton have a hierarchical relationship; the position data is analyzed to obtain current relative position information, and the current relative position information is used for indicating the current relative position of the virtual character between the first target skeleton and the second target skeleton under the current posture; determining a relative attitude relationship between the current relative position information and target relative position information, wherein the target relative position information is used for indicating a target relative position between a first target skeleton and a second target skeleton of the virtual character under a target attitude; and adjusting the posture of the virtual character according to the relative posture relationship. Compared with the prior art, the attitude adjustment efficiency and accuracy can be improved.

This application provides a cross-subject EEG emotion recognition method based on similarity-based dynamic cue routing, belonging to the field of EEG emotion recognition technology. The method includes: acquiring target domain samples and multiple source domain samples formed from subject EEG data and mapping them to target shared latent representation and multiple source domain shared latent representations, respectively; injecting corresponding source domain cue vectors into the source domain shared latent representation of each source domain to obtain enhanced source domain feature representations of multiple source domains; calculating the similarity between each source domain cue vector and the target shared latent representation, converting each vector similarity into sample-level source domain routing weights, and weighting and combining each source domain cue vector to generate dynamic cue; injecting the dynamic cue into the target shared latent representation to obtain enhanced target feature representations; and inputting the enhanced source domain feature representations and enhanced target feature representations of multiple source domains into a multi-branch neural network to train and construct an emotion recognition model.

The present invention provides a manually operated diaphragm valve that maintains compactness through a simple construction while simultaneously allowing for setting the valve body's opening and closing range. This design is unaffected by individual differences in the valve's internal shape, ensuring sufficient flow rate when fully open, and also allowing for arbitrary adjustment of the fully open flow rate. The valve body 1 includes a valve body 12 assembled inside a main body 10 and a main shaft 13 threaded into the main body 10 in a lifting motion, separating the valve body 12 from the valve seat 14. A handle 2 is mounted above the valve body 1 for manually rotating the main shaft 13. Near the upper part of the valve body 1, a fully open position adjustment mechanism 20 is provided. This mechanism restricts the rotation of the handle 2 on the fully open side, setting the upper limit of the flow rate of the valve body 12 when fully open.

The application relates to the technical field of fault early warning, and particularly discloses a bag making machine fault risk early warning method based on visual detection, which comprises the following steps: step S1: determining a monitoring point on a fault-free bag making machine sample, extracting a displacement sequence of the monitoring point, and generating a reference displacement period curve of the monitoring point; step S2: for a target bag making machine, real-time collection of working images of each period is carried out at a same preset frequency within several continuous working periods, a first displacement period curve of the monitoring point in the current working period is acquired, and a second displacement period curve of the next working period is synchronously acquired; step S3: first similarity of the first displacement period curve and the reference curve and second similarity of the first displacement and the second displacement are respectively calculated; variation curves of the two similarities are constructed within the continuous periods, the slopes of the two curves are calculated within the latest time window, and whether the target bag making machine has a fault risk is determined by analyzing the slope characteristics.

The invention provides a memory recognition method and device based on a game scene, and the method comprises the steps: achieving non-invasive physiological signal collection through a preset wearable device, reducing the wearing difficulty, and maintaining the immersion experience; acquiring baseline index data based on the memory baseline task to serve as a calibration value, and eliminating individual differences of users; calculating an electroencephalogram memory index score based on the obtained real-time electroencephalogram data and baseline index data, calculating a cognitive index score based on the obtained real-time behavior data, fusing the electroencephalogram memory index score and the cognitive index score to obtain a comprehensive memory cognitive index score, and performing memory recognition on the comprehensive memory cognitive index score. Therefore, the memory fluctuation of the player in the real-time game process can be accurately reflected; and the display time of the game main body is adjusted by using the comprehensive memory cognition index score, so that the task difficulty is balanced and optimized in real time, the hysteresis quality and insufficient individual response in the traditional method are effectively avoided, the heart stream experience of players is remarkably enhanced, and the interaction and immersion are improved while the scientific evaluation value is ensured.

The application provides a respiratory flow signal processing method based on dynamic calibration and normalization, individual differences are eliminated through individual treatment course calibration, temperature and humidity interference is eliminated through environment adaptive preprocessing, and zero point calibration precision is improved through adaptive Kalman filtering, so that the quantization precision (error is less than or equal to 5%) and cross-device consistency (error is less than or equal to 3%) of the respiratory flow signal are significantly improved. The method can be widely applied to the respiratory rehabilitation training scenes of children, the elderly and adults, and provides accurate signal input for VR respiratory interaction.

The application provides a rust removal area identification and area accurate calculation method for a curved surface workpiece, comprising: acquiring a surface image of a cylindrical workpiece, and establishing a calibration relationship between image pixel size and actual physical size; selecting a to-be-detected area and generating a corresponding binary mask; mapping pixel coordinates in the to-be-detected area to a cylindrical parameter space; according to the cylindrical geometric relationship and the mapping relationship, calculating an area correction weight for each pixel in the to-be-detected area, the weight being used for correcting area representation errors of the pixel caused by the cylindrical curvature and imaging projection, and the correction weights of all the pixels forming a curved surface area weight matrix; converting the image to a LAB color space, identifying a metal base color area and generating a rust removal candidate area; removing rust and oxide skin areas in the rust removal candidate area, and generating a target area mask after morphological optimization; and performing pixel-by-pixel weighted summation on the target area mask and the curved surface area weight matrix to obtain an actual physical area of the rust removal area.

This invention relates to the field of nondestructive testing technology for magnetic materials, specifically to a nondestructive testing method for the magnetic properties of ferrite magnets. This invention constructs a dynamic calibration relationship and a comprehensive magnetic property damage index, combined with an automated testing process, to solve the problem of rapid, online, nondestructive, and comprehensive inspection of ferrite magnets. A synchronous belt conveyor mechanism allows the magnet under test to pass directly through a ring-shaped testing channel. A pulsed magnetic field generator and an array of Hall sensors are used for in-situ saturation magnetization and magnetic field acquisition, eliminating the need to damage the sample and achieving truly online comprehensive inspection. Magnetic field data from at least three locations are collected, fused, and quantitatively evaluated to assess key magnetic property parameters. Magnet model information is entered, a dynamic calibration relationship is established using standard samples, and a comprehensive magnetic property damage index is calculated using preset weighting coefficients. Precise judgment is made based on preset thresholds, providing a rapid, accurate, and adaptable nondestructive testing solution for complex ferrite magnets.