Usage and health-triggered machine response
Systems and methods are provided that capture a user's physically observable state, such as body position. A processor analyzes the image to determine if the user's state is unhealthy and, if so whether or not the user is in the unhealthy state for longer than a threshold period of time. If the user is in an unhealthy state, the processor signals an alerting component to energize a circuit. Accordingly, the user may be prompted to alter their physical state to a less unhealthy more healthful state.
1. A system, comprising:
an imaging component to capture an image of a user while engaged with an electronic device and provide an imager output corresponding to the image;
a processor to receive the imager output and determine therefrom whether a physical state of the user matches an unhealthy physical state, wherein upon determining that the physical state matches the unhealthy physical state, the processor provides an output signal, wherein the unhealthy physical state comprises an unhealthy posture, wherein the unhealthy physical state comprises the unhealthy posture associated with at least one of muscle strain or joint strain, and wherein the processor determining whether the physical state of the user matches the unhealthy physical state comprises providing the image to a neural network trained to determine unhealthy physical states and receiving a health-indicating decision therefrom; and
an alerting component comprising an electronic circuit, wherein the alerting component, upon receiving the output signal, energizes the electronic circuit; and
wherein the neural network is trained, comprising:
collecting a set of user images;
applying one or more transformations to each user image of the set of user images, the transformations including mirroring, rotating, smoothing, sharpening, altering contrast, altering brightness, and altering color saturation to create a modified set of user images;
creating a first training set comprising the collected set of user images, the modified set of user images, and a set of non-unhealthy user images;
training the neural network in a first training stage using the first training set; and
creating a second training set for a second training stage comprising the first training set and the set of non-unhealthy user images that are incorrectly detected as unhealthy user images after the first training stage; and
training the neural network in the second training stage using the second training set.
2. The system of claim 1 , wherein the unhealthy physical state comprises a posture.
3. The system of claim 1 , wherein the unhealthy physical state comprises an absence of spectacles.
4. The system of claim 1 , wherein:
the imaging component to capture the image comprises a series of images of the user while engaged with the electronic device and provides the imager output corresponding to the series of images; and
the processor receives the imager output and determines therefrom that the physical state of the user matches the unhealthy physical state for a duration of time in excess of a previously determined period of time, wherein the duration of time is further determined by the processor utilizing the imager output corresponding to the series of images and providing the output signal in response thereto.
5. The system of claim 4 , wherein:
the electronic device comprises the electronic circuit further comprising a first portion and a second portion;
the processor receives a first image of the series of images, and a second image of the series of images, wherein the second image is received after the first image;
the processor, in response to receiving the first image, determines therefrom that the physical state of the user matches the unhealthy physical state and provides the output signal comprising a first output signal;
the processor, in response to receiving the second image, determines therefrom that the physical state of the user matches the unhealthy physical state and provides the output signal comprising a second output signal;
the alerting component, upon receiving the first output signal, energizes the first portion; and
the alerting component, upon receiving the second output signal, energizes the second portion.
6. The system of claim 1 , wherein the alerting component comprises a display device and the electronic circuit comprises a graphical element generator.
7. The system of claim 1 , wherein the alerting component comprises a physical actuator determining at least a portion of the user and the electronic circuit comprises a position controller of the physical actuator.
8. The system of claim 1 , wherein the alerting component comprises a networked component and the electronic circuit comprises an output portion of the networked component.
9. A method comprising:
capturing, by an imaging component, an image of a user while engaged with an electronic device and providing an imager output corresponding to the image;
receiving, by a processor, the imager output and the processor further determining therefrom whether a physical state of the user matches an unhealthy physical state, wherein upon determining that the physical state matches the unhealthy physical state comprising an unhealthy posture associated with at least one of muscle strain or joint strain, the processor provides an output signal, wherein determining whether the physical state of the user matches the unhealthy physical state comprises providing the image to a neural network trained to determine unhealthy physical states and receiving a health-indicating decision therefrom; and
receiving, by an alerting component comprising an electronic circuit, the output signal and, in response, energizing the electronic circuit; and
wherein the neural network is trained, comprising:
collecting a set of user images;
applying one or more transformations to each user image of the set of user images, the transformations including mirroring, rotating, smoothing, sharpening, altering contrast, altering brightness, and altering color saturation to create a modified set of user images;
creating a first training set comprising the collected set of user images, the modified set of user images, and a set of non-unhealthy user images;
training the neural network in a first training stage using the first training set;
creating a second training set for a second training stage comprising the first training set and the set of non-unhealthy user images that are incorrectly detected as unhealthy user images after the first training stage; and
training the neural network in the second training stage using the second training set.
10. The method of claim 9 , wherein the unhealthy physical state comprises a posture.
11. The method of claim 9 , wherein the unhealthy physical state comprises an absence of spectacles.
12. The method of claim 9 , further comprising:
capturing, by the imaging component, the image comprising a series of images of the user while engaged with the electronic device and providing the imager output corresponding to the series of images; and
receiving, by the processor, the imager output and determining therefrom that the physical state of the user matches the unhealthy physical state for a duration of time in excess of a previously determined period of time, further comprising determining the duration of time utilizing the imager output corresponding to the series of images and providing the output signal in response thereto.
13. The method of claim 12 , wherein:
the electronic device comprises the electronic circuit further comprising a first portion and a second portion;
receiving, by the processor, further comprises receiving a first image of the series of images, and a second image of the series of images, wherein the second image is received after the first image;
providing, by the processor, further comprises, upon receiving the first image, determining therefrom that the physical state of the user matches the unhealthy physical state, the output signal comprising a first output signal;
providing, by the processor, further comprises, upon receiving the second image, determining therefrom that the physical state of the user matches the unhealthy physical state and providing the output signal comprising a second output signal;
energizing the first portion, by the alerting component, upon receiving the first output signal; and
energizing the second portion, by the alerting component, upon receiving the second output signal.
14. The method of claim 9 , wherein the alerting component comprises a display device and the electronic circuit comprises a graphical element generator.
15. The method of claim 9 , wherein the alerting component comprises a networked component and the electronic circuit comprises an output portion of the networked component.
16. A system, comprising:
means to capture an image of a user while engaged with an electronic device and provide an imager output corresponding to the image;
means to receive the imager output and determine therefrom whether a physical state of the user matches an unhealthy physical state comprising an unhealthy posture associated with at least one of muscle strain or joint strain, and upon determining that the physical state matches the unhealthy physical state, means to provide an output signal, wherein the means to determine whether the physical state of the user matches the unhealthy physical state comprises means to provide the image to a neural network trained to determine unhealthy physical states and receiving a health-indicating decision therefrom; and
means to receive the output signal and, in response, energize an electronic circuit of an alerting component; and
wherein the neural network is trained, comprising:
collecting a set of user images;
applying one or more transformations to each user image of the set of user images, the transformations including mirroring, rotating, smoothing, sharpening, altering contrast, altering brightness, and altering color saturation to create a modified set of user images;
creating a first training set comprising the collected set of user images, the modified set of user images, and a set of non-unhealthy user images;
training the neural network in a first training stage using the first training set;
creating a second training set for a second training stage comprising the first training set and the set of non-unhealthy user images that are incorrectly detected as unhealthy user images after the first training stage; and
training the neural network in the second training stage using the second training set.
17. The system of claim 16 , wherein the unhealthy physical state comprises a posture.
18. The system of claim 16 , wherein the unhealthy physical state comprises an absence of spectacles.
19. The system of claim 16 , wherein:
the means to capture the image comprising a series of images of the user while engaged with the electronic device and provide the imager output corresponding to the series of images; and
the means to receive the imager output and determine therefrom that the physical state of the user matches the unhealthy physical state for a duration of time in excess of a previously determined period of time, wherein the means to determine the duration of time further comprises means to utilize the series of images and provide the output signal in response thereto.
20. The system of claim 16 , wherein the alerting component comprises a display device and the electronic circuit comprises a graphical element generator.