181results about How to "Good explainability" patented technology

A system and method for software estimation. In one embodiment, the software estimation system comprises a pre-processing neuro-fuzzy inference system used to resolve the effect of dependencies among contributing factors to produce adjusted rating values for the contributing factors, a neuro-fuzzy bank used to calibrate the contributing factors by mapping the adjusted rating values for the contributing factors to generate corresponding numerical parameter values, and a module that applies an algorithmic model (e.g. COCOMO) to produce one or more software output metrics.

The invention relates to a method of longitudinal control of curve driving of an autonomous vehicle. The method includes: judging the stage of the vehicle in the curve according to the vehicle state and driving route information; according to the stage of the vehicle in the curve, performing online real-time control on the driving speed of the vehicle and transferring control results to an acceleration control module; when the vehicle is judged to be in the state of driving inside the curve, calculating the difference value of the current speed and a driver's comfortable speed under the current curvature and transferring the difference value as expected acceleration to a lower acceleration tracking module for real-time control; when the vehicle is judged to be in the state of driving into the curve or out of the curve, outputting the expected acceleration in real time according to a driver model obtained by training and transferring the expected acceleration to the lower acceleration tracking module for real-time control. Driving characteristics of the singular driver are fully considered, the driving characteristics of the driver driving in the curve can be effectively simulated by controlling the demonstrated control characteristics in real time, and driver's acceptance of an automatic driving technique is improved.

The invention discloses a two-stage network image text cross-media retrieval method, which comprises the following steps of: firstly, exploring two-stage alignment by constructing a cross-media two-stage network, which respectively comprises two subnets for global and local; and training the cross-media two-stage model by using the training data set to determine network parameters in the cross-media two-stage model, thereby obtaining the trained cross-media two-stage mode; And finally, carrying out similarity retrieval on the to-be-retrieved image and the to-be-retrieved text by using the trained cross-media two-stage model. Experiments show that the cross-media retrieval method achieves a good effect in the application of cross-media retrieval.

A system and method for software estimation. In one embodiment, the software estimation system comprises a pre-processing neuro-fuzzy inference system used to resolve the effect of dependencies among contributing factors to produce adjusted rating values for the contributing factors, a neuro-fuzzy bank used to calibrate the contributing factors by mapping the adjusted rating values for the contributing factors to generate corresponding numerical parameter values, and a module that applies an algorithmic model (e.g. COCOMO) to produce one or more software output metrics.

The invention discloses a high-spatial-resolution pumping-detection micro-zone measurement device, system and method. The device comprises a laser device, a light splitter, a pumping laser adjusting assembly, a detection laser adjusting assembly, a beam combiner, a micro-zone measurement assembly and a light intensity detection assembly, wherein the pumping laser adjusting assembly is used for adjusting one of two beams of laser to obtain pumping laser; the detection laser adjusting assembly is used for adjusting the other one of the two beams of laser to obtain detection laser; the beam combiner is used for carrying out beam combination on the pumping laser and the detection laser; the micro-zone measurement assembly is used for positioning a measurement position of a sample to be detected in a microscopic manner before measurement, focusing the pumping laser and the detection laser and detecting the sample to be detected; the light intensity detection assembly is used for detecting the light intensity of the detection laser and the light intensity of transmission laser of the sample to be detected. By adopting the high-spatial-resolution pumping-detection micro-zone measurement device, system and method, a high-spatial-resolution micro-zone optical image and a detection image of the sample to be detected can be provided; the high-spatial-resolution pumping-detection micro-zone measurement device, system and method have the advantages of good imaging effect, good linear effect and the like.

The invention discloses a human movement detection method based on movement dictionary learning. The human movement detection method includes that at the training stage, using a local property representing method to extract human movement properties from different video clips, and learning a human movement dictionary with strong distinguishing ability through training; considering reconstructing errors and new errors when modeling the movement dictionary so as to model better; at the testing stage, enabling a space-time sliding window to traverse the sparse codes of sliding windows of the whole video, and judging whether the space-time sliding window comprises a human movement according to the response values of the sparse codes to different dictionary items. The human movement detection method based on the movement dictionary learning can obtain the human movement dictionary through training without a negative sample, and the training process is easy and quick to finish.

The invention discloses a gene expression time series data classification method based on a visibility graph algorithm. The method comprises the steps of (1) constructing a basic network, selecting adata strip according to preprocessed gene expression time series data, constructing a visibility graph and a connection graph by using the visibility graph algorithm, and determining a basic structureof a co-expression network, (2) extracting relevant traditional features according to the obtained basic network, (3) obtaining a feature vector of each gene node in the basic network by using second-order random walk and neural network model learning, and (4) integrating features of the basic network and using different strategies based on the obtained features of the basic network through a density clustering algorithm to complete the classification of gene expression time series data. The invention provides a method for realizing the gene expression time series data classification by usingvisibility graph foundation network construction, node feature vector extraction and the density clustering algorithm, and the method has good precision and practical performance.