Google Looking for Users’ Help to Improve Their Image Recognition Algorithms
Considering the colossal amounts of data Google has amassed over the years, it isn’t unusual for the company for put that data to enhance their machine learning capabilities. Thankfully, unlike another website that has ‘acquired’ copious amounts of user data, Google actually uses it for some good. Features such as the Google Assistant, Google Lens, Smart Replies etc. are a product of extensive machine learning algorithms, and they’re actually useful. Google is constantly working to improve the image recognition algorithms, and even made some of it open-source, for the world to see.
As weird as it may sound, the company’s AI systems still lack data for emerging markets such as India. Google is hoping they can get more human input to fix this. Currently, the company’s AI systems work pretty well in parts of the world such as the U.S, Europe etc. However, the results aren’t as impressive elsewhere around the world. Anurag Batra, product manager at Google, says “We have a very sparse training data set from parts of the world that are not the United States and Western Europe.” Batra is in charge of a project that aims to improve this situation. His team has built an app that asks people to perform easy tasks while simultaneously checking the accuracy of the company’s image recognition and translation algorithms.
To remedy the situation Google is promoting the Crowdsource app in India and other parts of Asia, but anyone around the world can download it. An expansion to Latin America is in the books too, later on in the year. Once you download the app, you’ll be asked to verify labels, help transcribe handwriting, help identify common household objects specific to a certain region and perform other similar tasks that help identify things. Anyone who has used Google Lens in India will know that objects are often mislabelled, very often ending in hilarious results.
For those looking to make a quick buck, no this isn’t like Google Opinion Rewards. There are no monetary rewards for user input. You’ll have to be content with points, badges, and certificates. Users who collect enough will be invited to join groups with other top contributors. If the prospect of identifying mundane objects for the greater good piques your interest, you can download the app using this link.