Appropriate modification of video call images

1. An apparatus comprising:

a video camera;

a processor;

a memory that stores code executable by the processor to:

determine a video image from the video camera satisfies a video policy for a video communication using a neural network trained with a clothing model by presenting images of users in clothing identified as one of appropriate and not appropriate to the neural network and with facial looks identified as desirable and undesirable, wherein the undesirable facial looks are identified in response to the user being made up and/or appearing ill, the neural network comprises input neurons, hidden neurons, and output neurons, the input neurons receive the images of the users in clothing identified as one of appropriate and not appropriate with the facial looks identified as desirable and undesirable, and the video policy is not satisfied in response to presenting the video image to the input neurons of the neural network to the output neurons of the neural network indicating a user in the video image is not appropriately dressed and has the undesirable facial look, and the video policy is further not satisfied in response to a user leaving the field of view of the video communication and in response to a third-party user that was not previously participating in the video communication enters a field of view of the video communication;

in response to the video image satisfying the video policy, present the video image for the video communication; and

in response to the video image not satisfying the video policy, obscure the user in the video image to conform to the video policy.

2. The apparatus of claim 1 , wherein the video image is further not presented to modify the video image to conform to the video policy.

3. The apparatus of claim 1 , wherein the code is further executable to:

determine an audio stream satisfies an audio policy for an audio communication;

in response to the audio stream satisfying the audio policy, provide the audio stream for the audio communication; and

in response to the audio stream not satisfying the audio policy, modify audio stream to conform to the video policy.

4. A method comprising:

determining, by use of a processor, a video image satisfies a video policy for a video communication using a neural network trained with a clothing model by presenting images of users in clothing identified as one of appropriate and not appropriate to the neural network and with facial looks identified as desirable and undesirable, wherein the undesirable facial looks are identified in response to the user being made up and/or appearing ill, the neural network comprises input neurons, hidden neurons, and output neurons, the input neurons receive the images of the users in clothing identified as one of appropriate and not appropriate with the facial looks identified as desirable and undesirable, and the video policy is not satisfied response to presenting the video image to the input neurons of the neural network to the output neurons of the neural network indicating a user in the video image is not appropriately dressed and has the undesirable facial look, and the video policy is further not satisfied in response to a user leaving the field of view of the video communication and in response to a third-party user that was not previously participating in the video communication enters a field of view of the video communication;

in response to the video image satisfying the video policy, presenting the video image for the video communication; and

in response to the video image not satisfying the video policy, obscuring the user in the video image to conform to the video policy.

5. A program product comprising a computer readable storage medium that stores code executable by a processor, the executable code comprising code to:

determine a video image satisfies a video policy for a video communication;

in response to the video image satisfying the video policy, present the video image for the video communication using a neural network trained with a clothing model by presenting images of users in clothing identified as one of appropriate and not appropriate to the neural network and with facial looks identified as desirable and undesirable, wherein the undesirable facial looks are identified in response to the user being made up and/or appearing ill, the neural network comprises input neurons, hidden neurons, and output neurons, the input neurons receive the images of the users in clothing identified as one of appropriate and not appropriate with the facial looks identified as desirable and undesirable, and the video policy is not satisfied in response to presenting the video image to the input neurons of the neural network to the output neurons of the neural network indicating a user in the video image is not appropriately dressed and has the undesirable facial look, and the video policy is further not satisfied in response to a user leaving the field of view of the video communication and in response to a third-party user that was not previously participating in the video communication enters a field of view of the video communication;

in response to the video image satisfying the video policy, present the video image for the video communication; and

in response to the video image not satisfying the video policy, obscure the user in the video image to conform to the video policy.

6. The program product of claim 5 , wherein the video image is further not presented to modify the video image to conform to the video policy.