What Does Microsoft Corporation Artificial Intelligence, Technology Boost Imply For Investors

Microsoft Corporation (NASDAQ:MSFT) recently started a competition to boost technology pertaining to artificial intelligence. The MS COCO Captioning Contest 2015 scheduled to end on Friday, has provided a platform for university technical specialists and professionals to unite and create the best algorithm that auto-captions pictures. Ultimately, the results of the competition would be announced in June at the CVR computer vision conference.


Microsoft on Thursday announced news pertaining to software development related to Artificial Intelligence and Technology (AIT). The software would create a link between language and visual, make captions for pictures and sort the image in a specific category. The technology is slightly similar to Google Inc. (NASDAQ:GOOG) Voice search technology and Siri by Apple Inc. (NASDAQ:AAPL).
The software works by breaking down that isentered by the user, and processed by separate servers to process and analyzes the information. The information is also processed through a build to understand the information in terms of patterns and common search results found, so the user gets the most accurate answer to the instructions s/he inputs.

According to researcher Xiaodong He, of Microsoft Research’s Deep Learning Technology Center (DLT) and his team, the machine is under training of Humanistic understanding of pictures. So far the machine has learnt the ability to hear, speak and understand almost the same way a human does. The team of researchers is also attempting to train the machine to focus on a human without having to input data of the background.
Talking about visual understanding, the company also stated that the technology would be able to recognize the subject in an image based on specific characteristics. The technology’s recognition software would be able to focus on the person sitting in front of a computer to be a man in the picture due to recognizing the human appearance and presence of facial hair.

Additionally Margaret Mitchell, a researcher at Microsoft Research’s National Language Processing Group (NLP) stated that technology has had great success due to the neural networking. Even though the software is not flawless, the idea behind auto-captioning by Microsoft would significantly impact the market.
Microsoft explained how the technology operates, stating that the process takes a number of steps. The system initializes a “convolutional neural network” which would recognize words that ideally describe the image. The visual part of the process recognizes colors and shapes while the language build correct the grammar of the chosen sequence of words. The technology also incorporates a last step, as a checker is posted to ensure the caption correlates to the picture.

The researchers from company concluded that as personal computers keep getting more intelligent, the technology would prove to be very applicable. They added that the extensive research would contribute to learning of technology, making it possible to create the most sophisticated robot capable of AI. The company’s optimism towards such a project is based on the creation of such technology where it only has to improve the system algorithm, eventually improving the technology’s understanding about world around it.

The plausible future of the technology could also incorporate videos letting users scan a subject and recognize it. The technology would also be able to understand the link between the instructions the user gives and the video being scanned at that moment. Furthermore, the technology’s application would be used in security cameras to help detecting humans and other possible subjects to pick project information that humans may have missed in the scene. Understanding this, the applications expand over a vast area from the health industry, photography hobbies to military purposes.

Source : bidnessetc