3 results about "Independent vector" patented technology

Methods and systems for image alignment are provided. In one embodiment, a method includes individually aligning candidate alignment target images in an image generated for a sample to corresponding set images of set alignment targets selected to have mutually linearly independent vectors between a position of the set alignment targets and a reference position. The method also includes, for any of the candidate images successfully aligned to their corresponding set images, individually determining coordinates of the reference position from coordinates of the aligned candidate images and their corresponding mutually linearly independent vectors. In addition, the method includes determining final coordinates of the reference position in the image generated for the sample from the individually determined coordinates of the reference position.

The invention relates to the technical field of artificial intelligence, in particular to a molecular representation distribution external generalization learning method based on annealing noise enhancement and dual vector quantization and application thereof. Extracting node characterization and invariance scores of the molecular graph by using a double-flow encoder to realize preliminary deentanglement of characterization; then, annealing noise enhancement mechanism mixed features are introduced, context dependence is reserved in the initial stage of training, two independent vector quantization experts are utilized to establish exclusive discrete potential spaces for invariant representation and false representation respectively, and the problem of information entanglement of a single codebook is solved; and meanwhile, node-level statistical independence is forcibly realized by utilizing an independence enhancement module, and joint optimization is carried out in combination with an invariant learning strategy. The objective of the invention is to solve the problem of how to realize thorough de-entanglement of invariant representation and false representation of non-Euclidean molecular graph data on a fine-grained node level.

Approaches presented herein provide systems and methods for determining duplicate objects within an interaction environment. Connectivity information for an object may be used to map a set of three linearly independent vectors corresponding to a transform applied to the object. These three linearly independent vectors may be used to form canonical forms of first and second objects to determine whether the first object and the second object are duplicates or near-duplicates. Copies of duplicate or near-duplicate objects may then be deleted from the interaction environment and represented by a common object to which one or more additional transforms are applied.