4results about How to "Guaranteed aggregation" patented technology

This application provides a method and apparatus for open-set skeleton action recognition based on outlier prototype learning. The method includes: constructing a neural network model; preprocessing human skeleton data by obtaining initial features through a multi-branch feature extraction network; processing the initial features using a classifier and a hypersphere feature mapper to obtain logical prediction values ​​for action categories and branch features; training the neural network model based on a training set; after the first iteration, optimizing the feature space of in-distribution samples using multi-class loss; synthesizing virtual outliers in the optimized feature space of in-distribution samples; optimizing the energy boundary by combining in-distribution samples and virtual outliers; after the second iteration, selecting the optimal model weights based on the combined performance of open-set recognition and closed-set classification on the validation set; and determining the action type and triggering the corresponding robot operation based on the energy score output by the trained neural network model and the action classification result.

The utility model provides a dilution device for solid food and drug detection, which relates to the technical field of food and drug detection and comprises a support frame, a center plate is fixedly mounted on the support frame, a swing frame positioned above the center plate is hinged to the support frame, and a hydraulic rod is fixedly mounted in the center of the upper surface of the swing frame in a penetrating manner. The grinding plate is arranged, solid food and medicine placed on the center plate are ground through descending of the grinding plate, the situation that the particle size is too small due to excessive grinding can be avoided through grinding, and therefore loss or instability of effective components of medicine is reduced; the uniformity of the medicine can be better controlled through grinding of the grinding plate, and dosage deviation caused by uneven grinding is avoided.

The application discloses a kind of bonding materials and its preparation method.Preparation method includes the following steps: in bearing appliance is arranged bearing carrier, the bearing carrier includes the first inert liquid layer in bottom layer 350CS~400CS, the second inert liquid layer in intermediate layer 200CS~300CS And the third inert liquid layer in top layer 50CS~100CS;By microfluidic process to generate several droplets on the bearing carrier with mixed solution as liquid phase, wherein the solute in the mixed solution includes 8wt%~10wt% Conductive powder, 8wt%~12wt% dispersing agent and 10wt%~12wt% binder;The bearing carrier containing the droplet is solidified and handled, and several microspheres are formed;And, the microspheres are calcined and handled.The application process is simple, low in cost and without solder residue problem.

This invention relates to the field of fire resistance testing of building components, and particularly to an intelligent temperature-controlled fire resistance testing device for building partition wall components. It includes a test assembly, with a compatible component installed on the inner wall of the test assembly, and a penetration detection component installed on the bottom of the test assembly. The invention activates a second electric push rod, moving a storage box to a suspected gap on the unexposed side of the specimen, pushing a cotton pad into the unexposed side of the specimen. Simultaneously, the inner cylinder descends, driving a gap probe to penetrate the cotton pad and reach the gap. Continuous compression further penetrates the gap probe and sliding tube into the inner cylinder. A scale structure is used to determine the crack depth. When a penetrating crack or hole appears, the cotton pad ignites, which is then monitored by a smoke sensor, thus improving the fire resistance testing effect of the partition wall components.