Cortexica Launches New Visual Search Engine

A new startup firm has developed a search engine that works just like the human eye. So you can use this technology with the help of smartphones to search for all the information regarding any object by clicking its photo. With the aid of this visual tool, you can take a snapshot of some object in a supermarket and get the online price listing of the product. In this way you can find where to get this product by spending the least amount of money. This sounds pretty cool. This biologically inspired search engine is developed by a startup called Cortexica and it was formed by the students while doing a research at Imperial College London. The company recently created an app called WINEfindr, used for comparing different wines and has proved to be a helpful tool for the wine tasters and critics. Then a week ago, this firm revealed its application-programming interface (API) for the technology, which will facilitate other web developers to make apps with similar functions.

#-Link-Snipped-#This technology is somewhat similar to those used by those bar code reading apps that can use the bar code information on any object to find its cost, manufacturing details and almost everything available about it on the internet. But the main advantage of this new app will be that it will directly recognize the object instead of the QR code on it. In a way, the phone camera will function like a human eye and search for it in the virtual world. Cortexica’s co-founder and researcher at the Imperial College, Anil Bharath said that the technology is inspired by the human visual system. It looks for certain key features and characteristics of an object to recognize it. It doesn’t rely much on the orientation, darkness or brightness, etc. to identify the object. This interesting twist in the tale helps the tool to identify even moving objects and far away products. The most awesome feature of this technology is that it can detect logos and objects even in videos.

The thing which makes Cortexica special is that it copies the edge extraction feature used by the human visual system. It helps it to compare the different points of an image with respect to its neighbors. This feature enables you to identify an object in varied conditions. T imitate this, Cortexica will use graphical processing units (GPUs) to solve the problem of parallel processing. Skeptics say that using variations and resolving them will complicate the computation and reliable matching will get tougher.

Although there are few other web services like Google’s Goggles, TinEye, etc. that are somewhat similar to this technology. However, people at Cortexica claim that the accuracy and precision provided by the new app is like nothing before. Time only can tell the truth in these claims.

Source: #-Link-Snipped-#