The standard theory of visual search tasks for visual search assumes that when a person searches for an object in an array of other items it uses memory to find the target. The analysis of the following three articles strongly supports this highly respected theory, and this theory shows that there may be some flaws in reasoning. In the article "Characteristics and Purpose in Visual Processing", Anne Treisman pointed out that there are two theoretical aspects to visual processing.
In the summer of 2016, with the introduction of object detection by Pinterest, a visual search has been developed and object detection can find all the objects in the image in real time and get appropriate results. Since that time, visual search has become one of the most popular features, detecting hundreds of millions of visual searches and billions of objects each month. Lens combines understanding and discovery techniques for images and objects of Pinterest to provide various results to Pinner. For example, if you take a picture of a blueberry, the lens will not only return blueberries. You can also cultivate beauty ideas such as blueberry scones, smoothies recipes, detox scrubs, and proprietary blueberry shrubs. skill
About a year ago, Pinterest introduced Lens to the world. This is a search tool that not only answers questions but also provides results through visual input that provides users with more rewards. This is a visual search tool that responds to all content and all things related to that input, not the same input picture. You ate a cake, you got a series of recipes, kapish? Then there is voice search. Audio AI is mature and has opportunities. Amazon and Google are exploring great success in this area as instruments like Alexa are building their place in our house. I realize that visual search will be the next most important, as the possibility of audio AI has been realized and accepted.
The future of the search is voice input or visual input. Text input is retained, but it may slowly disappear. Start searching for queries using voice or visual as input to the search engine, but the output will still be initially still on the network, mobile screen, home or office wall, or on the same TV or computer, the same sound There are also cases.