Quantcast

Google Speeds Up Web Search With Voice, Image Recognition

June 15, 2011

Google unveiled an array of new search features on Tuesday designed to accelerate and simplify Web searches as users increasingly access the Internet from handheld mobile devices.

Among the new features are voice and image recognition tools that allow users to use speech or images to conduct their searches more quickly.

At a briefing on Tuesday, members of Google’s search team unveiled the company’s latest improvements, saying they were designed to deliver information “in the blink of an eye.”

One of the new features, called “ËœInstant Pages’, predicts which link a searcher is likely to chose, and begins retrieving that page — even before the user clicks the link. 

Instant Pages utilizes a user’s Google search history, the relevance of each search result and roughly 200 other algorithmic factors to determine which link a user is most likely to click on.  It then begins contacting the server to load that page in the background as soon as the search results appear.

“We’re obsessed with speed,” said Google fellow and search scientist Amit Singhal at the “Inside Search” event in Mountain View, California.

“We call speed the killer app. None of us have enough time, and last year’s Google Instant was one of the biggest improvements we’ve made in getting information to users quicker,” he said.

Other enhancements include reducing the time it takes for Web pages to display after links are clicked on in search results.

“We at Google will not be happy until we make the Web as easy to flip through as a magazine,” he added.

“We measure every millisecond.”

On average, users spend 9 seconds entering a query and 15 seconds sifting through results, while Google’s processing between the two only takes 1 second, Singhal said.

“The time it takes Google to return a result is negligible compared to how long it takes the user to enter the query,” said Singhal, adding that it takes, on average, five seconds for a Web page to load once a user has clicked on a link in the search results.

The new features unveiled on Tuesday are spread across all of Google’s gadgets — from desktops using Chrome software to the latest Android-powered smartphones and tablet computers.

“In mobile, we are always thinking about how we can make the process of getting those results easier,” Google’s mobile engineering director Scott Huffman told the AFP news agency.

The new mobile search features include added icons to the bottom of search pages that allow users to conduct common searches with a single click, rather than having to type in specific queries.

Other mobile enhancements include the ability to build searches with simple “plus” buttons, and instant previews of search results that are viewable with simple swipes of a touchscreen.

Huffman said the “ËœGoogle Goggles’ feature, which lets mobile users with search using pictures, now translates languages in photos of text ““ with Russian having now been added to the list.

Johanna Wright, Google’s director of search product management, said the company was applying mobile innovations to desktop computers with the addition of voice and image search capabilities.

“Mobile has opened a world of possibilities,” Wright said.

Singhal said Google’s voice search, which allows mobile phone users to speak their search query, would be extended to desktop searches. 

The feature will be accessible for Chrome and Firefox users later this week, via a small microphone symbol next to the Google search box. The feature will initially be in English only, but will be extended to other languages later.

The new image search feature will allow users to upload, drag and drop, or paste a URL or use a browser extension to provide a reference to an image.  Google will then identify the image by comparing it to its database and publishing the subject, location and details, the Guardian reported.

However, Singhal said the facial recognition was not part of the image search process, and that all images shared during searches are anonymized, and not publicly accessible.

“If you uploaded a picture of John Lennon, it might find other similar pictures of John Lennon, but it wouldn’t know it was John Lennon,” the paper quoted him as saying.

“It’s an engine for analyzing color, lines and texture, not for facial recognition.”

Shares of Google’s stock rose $3.64, or 0.72 percent, on Tuesday, closing at $508.37.

On the Net:




comments powered by Disqus