Device setting adjustment based on content recognition

Abstract

Approaches provide for adjustment of playback settings on a user device based at least in part on a type of content being consumed via the user device. The content type may be determined based on a fingerprint acquired from the content presented via on the client device. The content type may be associated with one or more associated playback settings based on properties of the content type, which may be stored in a content profile. The playback settings may be provided to the user device to set or adjust one or more settings of the user device, as may relate to the presentation of the content.

1 .- 14 . (canceled) 
     
     
         15 . A computer-implemented method, comprising:
 receiving a fingerprint from a client device corresponding to at least a portion of rendered content rendered by the client device;   determining a type of the rendered content based at least in part on the received fingerprint;   evaluating one or more device settings that are associated with how the rendered content is provided;   determining one or more adjustments to the one or more device settings on the client device based on the determined type, the one or more adjustments changing how the rendered content is provided; and   for the one or more adjustments exceeding a threshold difference compared to the one or more device settings, applying the one or more adjustments to the provided rendered content for incremental application,   wherein the determined content type is transmitted to a neural network that:   evaluates, the machine learning system configured to evaluate the one or more device settings for a respective content type of the rendered content;   evaluates one or more content settings corresponding to rendering a particular content type; and   determines one or more settings consistent across the particular content type.   
     
     
         16 . The computer-implemented method of  claim 15 , further comprising:
 changing the one or more device settings based on the one or more adjustments that modifies at least one of the visual or auditory rendering of the content.   
     
     
         17 . The computer-implemented method of  claim 15 , further comprising:
 retrieving a stored type profile associated with the client device, the stored type profile including at least one client device setting for rendering content having a corresponding type of the rendered content; and updating the at least one client device setting of the stored type profile with the one or more adjustments.   
     
     
         18 . The computer-implemented method of  claim 15 , further comprising:
 retrieving a stored type profile associated with a client device, the stored type profile including at least one client device setting for rendering content having a corresponding type; and activating the stored type profile based on the determined type.   
     
     
         19 . The computer-implemented method of  claim 15 , further comprising:
 activating a second client device associated with the first client device, the second client device rendering at least a portion of the content; and adjusting one or more second client device settings based on the determined type.   
     
     
         20 . The computer-implemented method of  claim 15 , further comprising:
 receiving a user profile associated with a particular content type, the user profile comprising device settings associated with the rendered content; and modifying the one or more adjustments based on the device settings when the type is the particular content type.   
     
     
         21 . The computer-implemented method of  claim 19 , further comprising:
 determining a configuration of a room including a client device; determining one or more adjustments based on the configuration; and applying the one or more adjustments to the device settings.   
     
     
         22 . A computer-implemented method, comprising:
 training a neural network to associate training data including received playback settings with content types, to generate a trained model;   determining, by the trained model, a recommended playback setting based on an input content type; and
 updating the trained neural network based on received feedback. 
   
     
     
         23 . The computer-implemented method of  claim 21 , wherein the training data is ground truth that is applied so that the trained model identifies the recommended playback setting based on the input content type. 
     
     
         24 . The computer-implemented method of  claim 21 , further comprising training the neural network to recognize the content type, and determine potential combinations of playback settings to associate with the recognized content type. 
     
     
         25 . The computer-implemented method of  claim 21 , wherein the trained model further incorporates a device setting or a feature. 
     
     
         26 . A computer-implemented method, comprising:
 receiving, at a server, at least one of an audio or visual data associated with content;   determining, by a machine learning module, a content type associated with the received audio or visual data;   determining, by the machine learning module, a recommended playback setting based on the determined content type;   comparing the recommended playback setting to a profile setting, and obtaining a difference between the recommended playback setting and the profile setting; and   for the difference within a threshold, generating an output signal that includes the recommended playback setting.   
     
     
         27 . The computer-implemented method of  claim 25 , wherein the output signal is provided to gradually adjust, over a period of time, a device setting based on the recommended playback setting. 
     
     
         28 . The computer-implemented method of  claim 25 , wherein the computer-implemented method is performed in a chipset or in a computer-readable medium as stored executable instructions. 
     
     
         29 . The computer-implemented method of  claim 25 , wherein the machine learning module is applied to automatic content recognition, and the machine learning module further determines the content type based on object recognition. 
     
     
         30 . The computer-implemented method of  claim 25 , wherein the machine learning module comprises a neural network, and the neural network is a regression model that provides, as the recommended playback setting, a value on a continuous range of values associated with a potential content type associated with the content, or a classification model that provides, as the playback setting, a discrete value associated with the content type. 
     
     
         31 . The computer-implemented method of  claim 25 , wherein the machine learning module adjusts the recommended playback setting based on a relationship between the content type and the profile setting. 
     
     
         32 . The computer-implemented method of  claim 25 , wherein the machine learning module applies facial recognition or character recognition to extract identifying information and match rendered content against a library to identify the content type. 
     
     
         33 . The computer-implemented method of  claim 25 , wherein the output signal is received by a device that provides at least one of a video output and an audio output.