Artificial intelligence apparatus for recognizing object and method therefor

A method of recognizing at least one or more objects, comprising:

• obtaining image data for the at least one or more objects;
• generating first identification information corresponding to the at least one or more objects from the image data using a default recognition model composed of at least one or more of the plurality of recognition models for generating identification information corresponding to the object from the image data;
• measuring a confidence level for generation of the first identification information;
• obtaining the first identification information as a recognition result of the at least one or more objects if the measured confidence level is equal to or greater than a first reference value;
• and
• obtaining the second identification information corresponding to the at least one or more objects from the image data as a recognition result of the object using a compound recognition model composed of at least one or more of the plurality of recognition models if the measured confidence level is less than a first reference value,
• wherein the default recognition model is a model defined by first weights for each of the plurality of recognition models,
• wherein the compound recognition model is a model defined by second weights for each of the plurality of recognition models,
• wherein the obtaining of the second identification information comprises:
• determining the second weights corresponding to each of the plurality of recognition models based on the image feature vector extracted from the image data and constructing the compound recognition model using the second weights based on the measured confidence level being less than the first reference value.

A machine-readable non-transitory medium having stored thereon machine-executable instructions for recognizing at least one or more objects, the instructions comprising:

• obtaining image data for the at least one or more objects;
• generating first identification information corresponding to the at least one or more objects from the image data using a default recognition model composed of at least one or more of the plurality of recognition models for generating identification information corresponding to the at least one or more objects from the image data;
• measuring a confidence level for generation of the first identification information;
• obtaining the first identification information as a recognition result of the at least one or more objects if the measured confidence level is equal to or greater than a first reference value;
• and
• obtaining the second identification information corresponding to the at least one or more objects from the image data as a recognition result of the at least one or more objects using a compound recognition model composed of one or more of the plurality of recognition models if the measured confidence level is less than a first reference value,
• wherein the default recognition model is a model defined by first weights for each of the plurality of recognition models,
• wherein the compound recognition model is a model defined by second weights for each of the plurality of recognition models,
• wherein the obtaining of the second identification information comprises:
• determining the second weights corresponding to each of the plurality of recognition models based on the image feature vector extracted from the image data and construct the compound recognition model using the second weights if the measured confidence level is less than the first reference value.