Deep-Learning image recognition is a hot-topic. The billion dollar thought is to create a “Google” of image search (or a mesh-search engine for 3D printing and animation), but that requires rather high search fidelity. A lower-fidelity approach is to use key information provided in selfies – specifically the identity of the individual in the picture – and use less precise image recognition capabilities to tag background information. The resulting key-value pairs (e.g. personId, tagged_background_item) are a valuable source of marketing information about the individual. With 1.8 billion selfies being posted every day, there is a very real chance that the noise in lower-fidelity image recognition techniques will cancel out leaving a decent probability successful tagging – and a resulting valuable marketing base.
Check out the TechEnablement Deep-Learning articles, IBM SyNAPSE chip technology, and TF/s to PF/s training tutorials accelerated by GPU and Intel Xeon Phi technology to see how you can “change the world” with this technology.
For more information: