Location tagging for visual data of places using deep learning

Abstract

A method, computer system, and a computer program product for generating a location tag for a piece of visual data using deep learning is provided. The present invention may include receiving the piece of visual data. The present invention may also include analyzing the received piece of visual data using a neural network. The present invention may then include retrieving a location for the analyzed piece of visual data from the neural network. The present invention may further include generating a plurality of metadata for the retrieved location associated with the analyzed piece of visual data, wherein the generated plurality of metadata includes the location tag.

What is claimed is: 
     
         1 . A method for generating a location tag for a piece of visual data using deep learning, the method comprising:
 receiving the piece of visual data from a software application associated with at least a camera, a mobile phone or a social media site, wherein each received piece of visual data is validated for veracity;   analyzing the received piece of visual data using a neural network, wherein the neural network is built by receiving a plurality of training visual data, wherein the received plurality of training visual data is analyzed using a plurality of nodes from the neural network, wherein the received plurality of training visual data is based on an authoritative source, wherein the authoritative source is pre-defined by the user;   wherein a piece of digitally altered visual data is identified; and   wherein a plurality of image data is determined based on the analyzed plurality of training visual data, wherein the determined plurality of image data is stored in a large labeled dataset within the neural network;   determining the received plurality of training visual data includes a plurality of corresponding location metadata, wherein the received plurality of training visual data is stored in the large labeled dataset within the neural network with the plurality of corresponding location metadata;   determining the received plurality of training visual data excludes a plurality of corresponding location metadata, wherein the received plurality of training visual data is analyzed using the neural network, wherein a plurality of locations corresponding with the received plurality of training visual data is received, wherein the received plurality of locations is associated with the received piece of training visual data; wherein the received plurality of locations with the associated received plurality of training visual data is stored in the large labeled dataset within the neural network;   determining the retrieved location corresponds with the received plurality of training visual data, wherein the user fails to select a manual annotation option to change the determined location retrieved by the neural network, wherein computer generated data associated with weather, computer generated data associated with time of day and computer generated data associated with date is utilized to confirm a genuineness of the analyzed piece of visual data, wherein the genuineness of the analyzed piece of visual data includes a determination on the accuracy of the retrieved location associated with the analyzed piece of visual data;   in response to a determination that the retrieved location fails to correspond with the retrieved plurality of training visual data and the analyzed piece of visual data lacks genuineness, failing to generate a plurality of metadata for a location for the analyzed piece of visual data;   retrieving a location for the analyzed piece of visual data from the neural network, wherein the large labeled dataset within the neural network is searched for a corresponding location to the analyzed piece of visual data, wherein the corresponding location for the analyzed piece of visual data is retrieved from the large labeled dataset within the neural network;   generating the plurality of metadata for the retrieved location associated with the analyzed piece of visual data, wherein the generated plurality of metadata includes the location tag;   presenting the retrieved corresponding location for the received piece of visual data to a user, wherein the retrieved corresponding location comprise a plurality of additional location-based information and genuineness of the piece of visual data;   receiving a revised location for the received piece of visual data, wherein the plurality of additional location-based information corresponds with the retrieved location for the received piece of visual data;   storing, in the large labeled dataset within the neural network, the revised location corresponding with the received piece of visual data; and   generating a revised plurality of metadata based on the revised location associated with the received piece of visual data.