Selective detection of visual cues for automated assistants

An assistant device that facilitates touch-free invocation of an automated assistant, comprising:

• one or more vision sensors;
• one or more processors;
• and
• memory storing instructions that, in response to execution of the instructions by the one or more processors, cause the one or more processors to:
• detect one or more stimuli;
• in response to the one or more stimuli, perform an excluded region classification process to classify one or more regions of a field of view of the vision sensor as being likely to contain visual noise and/or unlikely to contain user-provided visual cues, wherein during the region classification process:
• one or more image frames are obtained from the vision sensor;
• the one or more image frames are analyzed;
• and
• based on the analysis, the one or more regions of the field of view of the vision sensor are classified as likely to contain visual noise;
• wherein as a result of the classification, visual cues detected in one or more of the classified regions are less likely to invoke the automated assistant than visual cues detected subsequently outside of the one or more classified regions, or the one or more classified regions are not analyzed for visual cues to invoke the automated assistant in subsequent image frames obtained from the vision sensor.

At least one non-transitory computer-readable medium comprising instructions that, in response to execution of the instructions by one or more processors of a vision-sensor equipped assistant device, cause the one or more processors to facilitate touch-free invocation of an automated assistant, including:

• detecting one or more stimuli;
• in response to the detecting, performing an excluded region classification process to classify one or more regions of a field of view of the vision sensor as being likely to contain visual noise and/or unlikely to contain user-provided visual cues, wherein the region classification process includes:
• obtaining one or more image frames from the vision sensor;
• analyzing the one or more image frames;
• and
• based on the analyzing, classifying the one or more regions of the field of view of the vision sensor as likely to contain visual noise;
• wherein as a result of the classifying, visual cues detected in one or more of the classified regions are less likely to invoke the automated assistant than visual cues detected subsequently outside of the one or more classified regions, or the one or more classified regions are not analyzed for visual cues to invoke the automated assistant in subsequent image frames obtained from the vision sensor.