Determining similarity of images using multidimensional hash vectors corresponding to the images

A computer system, comprising:

• one or more processors;
• memory;
• and
• one or more programs stored in the memory and configured for execution by the one or more processors, the one or more programs comprising instructions for:
• receiving an input image having content;
• generating a feature vector corresponding to the input image according to a trained classifier model, the feature vector having a plurality of components;
• encoding the feature vector as a similarity hash by quantizing each component;
• computing one or more additional hashes of the feature vector using additional hashing techniques, each additional hash generated by a respective distinct hashing technique;
• updating the similarity hash as a combination of the similarity hash and the one or more additional hashes;
• and
• for each reference image in a plurality of reference images:
• obtaining a reference hash for the respective reference image;
• computing similarity between the input image and the respective reference image by computing a distance between the reference hash and the similarity hash;
• determining if the computed distance is within a predetermined threshold;
• and
• in accordance with a determination that the computed distance is within the predetermined threshold, returning the respective reference image as an image that is similar to the input image.

A non-transitory computer readable storage medium storing one or more programs configured for execution by one or more processors of a computer system, the one or more programs comprising instructions for:

• receiving an input image having content;
• generating a feature vector corresponding to the input image according to a trained classifier model, the feature vector having a plurality of components;
• encoding the feature vector as a similarity hash by quantizing each component;
• computing one or more additional hashes of the feature vector using additional hashing techniques;
• updating the similarity hash as a combination of the similarity hash and the one or more additional hashes without quantizing the one or more additional hashes;
• and
• for each reference image in a plurality of reference images:
• obtaining a reference hash for the respective reference image;
• computing similarity between the input image and the respective reference image by computing a distance between the reference hash and the similarity hash;
• determining if the computed distance is within a predetermined threshold;
• and
• in accordance with a determination that the computed distance is within the predetermined threshold, returning the respective reference image as an image that is similar to the input image.