Automatic extraction of salient objects in virtual environments for object modification and transmission

A system comprising:

• at least one processor;
• and
• a memory coupled to the at least one processor, with software instructions stored thereon that, when executed by the at least one processor, cause the at least one processor to perform operations including:
• determining a reference avatar in a virtual environment;
• obtaining a plurality of properties of an object in the virtual environment, wherein the properties include at least one of:
• spatial properties within the virtual environment, visual properties, or audio properties;
• determining a plurality of saliency factors of the object, wherein each saliency factor is normalized to a same numeric range and is based on a different set of one or more properties of the plurality of properties of the object, and wherein one or more of the saliency factors are additionally based on one or more properties of the reference avatar;
• determining a saliency measure of the object with respect to the reference avatar based on a combination of the plurality of saliency factors, wherein the saliency measure is determined based on a weighted linear combination of the plurality of saliency factors, wherein at least one of the plurality of saliency factors is modified by a weight associated with the reference avatar or a weight associated with the virtual environment;
• determining that the saliency measure is greater than a threshold saliency measure associated with the reference avatar;
• in response to determining that the saliency measure is greater than the threshold saliency measure associated with the reference avatar,
• automatically modifying, without user input, at least one of the reference avatar or the object in the virtual environment;
• and
• in response to determining that the saliency measure is less than the threshold saliency measure associated with the reference avatar, omitting the modifying of the at least one of the reference avatar or the object.

A non-transitory computer-readable medium with instructions stored thereon that, when executed by a processor, cause the processor to perform operations comprising:

• determining a particular reference avatar in a virtual environment, wherein the particular reference avatar is one of a plurality of reference avatars associated with the virtual environment;
• obtaining a plurality of properties of an object in the virtual environment, wherein the properties include at least one of:
• spatial properties within the virtual environment, visual properties, or audio properties;
• determining a plurality of saliency factors of the object, wherein each saliency factor is normalized to a same numeric range and is based on a different set of one or more properties of the plurality of properties of the object, and wherein one or more of the saliency factors are additionally based on one or more properties of the particular reference avatar;
• determining a saliency measure of the object with respect to the particular reference avatar based on a combination of the plurality of saliency factors;
• determining whether the saliency measure is greater than a threshold saliency measure associated with the particular reference avatar, wherein the threshold saliency measure is based on particular settings associated with the particular reference avatar, and wherein one or more other reference avatars of the plurality of reference avatars are associated with respective threshold salience measures based on respective settings associated with each of the other reference avatars;
• in response to determining that the saliency measure is greater than the threshold saliency measure associated with the particular reference avatar, automatically modifying, without user input, at least one of the particular reference avatar or the object in the virtual environment;
• and
• in response to determining that the saliency measure is less than the threshold saliency measure associated with the particular reference avatar, omitting the modification of the at least one of the particular reference avatar or the object.