The goal of
this research is to determine a teachable machine’s ability to differentiate
the vocalizations of Canis lupus familiari. The research will begin with
inputting audio files of Canis lupus familiari vocalizing during
different events and varying environments. If the machine can determine the
reasons for the unique vocalizations, it should be able to output the reasons
for the various sounds, enabling humans to understand their canine companion more
effectively. The significance of the research is for humans to better
understand why Canis lupus familiari makes unique vocalizations, resulting
in improved communication between humans and Canis lupus familiari and
ultimately better care for the animal. This technology could be used not only
for a pet owner's communication with their animal but also for scenarios in
which the owner of the animal is away and a pet sitter is unsure of the reason behind
the vocalization. With the increased spotlight on the value of service animals,
a scenario where this technology could be useful is with people with social
communication disorder (SCD), bridging the gap between human and animal, and ensuring
the animal is cared for appropriately and the human knows when the service
animal is alerting for medical reasons.
The research will be conducted by
observing and recording Canis lupus familiari vocalizing in various
environments, triggered by a range of events. Each audio file will be notated
with the date and location of the event as well as the reason for the
vocalization. The audio files will then be stored, organized, labeled, converted,
and uploaded into a teachable machine for analysis, testing, and training.
Variable Table
|
Name
|
I/D/C
|
Description
|
|
Vocalization
|
D
|
Sound emitted from Canis lupus
familiari
|
|
Environment
|
C
|
Inside or outside
|
|
Triggering Event
|
I
|
Reason for vocalization
|
|
Background noise inside
|
C
|
Same House
|
|
Background noise outside
|
C
|
Same yard
|
Is
it possible for a teachable machine to interpret the reason Canis lupus
familiari is vocalizing based on a variety of audio files? It is believed
that a teachable machine will be able to determine and output the reason for Canis
lupus familiari vocalization and provide a percentage indicating the
confidence level of the model with its prediction. It may be discovered during the
research, that the teachable machine can recognize additional variations in vocalizations
for the same events and environments. There is a chance the teachable machine cannot
differentiate the sounds depending on the environment or the event.
