A Tongue-Tracking Artificial Larynx

Researchers in South Africa are working on a new kind of artificial larynx that won't have the raspy voice of existing devices. The system tracks contact between the tongue and palate to determine which word is being mouthed, and uses a speech synthesizer to generate sounds.

According to the National Cancer Institute, some 10,000 Americans are diagnosed with laryngeal cancer each year, and most patients with advanced cancer must have their voice box removed.

"All of the currently available devices produce such bad sound--it either sounds robotic or has a gruff speaking voice," says Megan Russell, a PhD candidate at the University of the Witwatersrand in Johannesburg, South Africa. "We felt the tech was there for an artificial synthesized voice solution."

The system uses a palatometer: a device that looks much like an orthodontic plate and is normally used for speech therapy. The device, made by CompleteSpeech of Orem, UT, tracks contact between the tongue and palate using 118 embedded touch sensors. The software for the artificial larynx was written by Russell and colleagues at the University of the Witwatersrand. Their work is being presented at the International Conference on Biomedical and Pharmaceutical Engineering this week in Singapore.

To use the device, a person puts the palatometer in her mouth and mouths words normally. The system tries to translate those mouth movements into words before reproducing them on a small sound synthesizer, perhaps tucked into a shirt pocket.

So far, Russell has trained the system to recognize 50 common English words by saying each word multiple times with the palatometer in her mouth. The information can be represented on a binary space-time graph and put into a database. Each time the user speaks, the contact patterns are compared against the database to identify the correct word.

Russell's team has tested the word-identification system using a variety of techniques. One approach involves aligning and averaging the data produced while training the device for a few instances of a word to create a template for comparison. Another compares features such as the area of the data plots on the graph, and the center of mass on the X and Y axes. A voting system compares the results of selected methods to see whether there is agreement. The researchers have also tested a predictive-analysis system, which considers the last word mouthed to help determine the next.

Russell says that when the voting and predictive elements are combined, the system identifies the correct word 94.14 percent of the time, although this doesn't include words that the system classifies as "unknown" and chooses to skip. Russell says that happens about 18 percent of the time. But choosing the wrong word "could lead to some very difficult social situations," Russell says, so it's best for the system to reject unclear words and remain silent.

Searching Facebook More Intimately

In the search industry's push to mine online social networks for improved results, the search engine Cuil has become the first to index information from your Facebook friends. Cuil then places direct and thematically related results from your Facebook network beside general Web search results.

The search offering, called Facebook Results, only works if you opt in from a Cuil search-return page. Once you do that, Cuil indexes your Facebook network in a few seconds. Afterward, any Cuil general Web search you perform also turns up items from your Facebook network and posts them in a right-hand column.

Cuil's search algorithms find direct and related results. For example, my search for "asthma" summoned Facebook posts from a friend who had started a health-care networking website, others from a high-school classmate writing about his cancer diagnosis (the word "diagnosis" was deemed relevant), as well as a few posts about people's colds and sinus complaints. A search for "Ecuador" turned up a travel agent acquaintance who was talking about a jungle tour, as well as a post from a journalist friend who was passing along a news story about the Congo (the technology picked up on the developing-nation theme).

In contrast, when I performed my "asthma" and "Ecuador" searches within Facebook, the Facebook engine gave me only general hits such as Facebook pages for asthma sufferers or national fan sites for Ecuador, but nothing at all from any of my friends' posts.

The Cuil technology is built on Facebook Connect, the existing Facebook interface that other websites use to gain exposure within the social network. Facebook permitted Cuil to indexes users' content--when permitted by individual users--on the condition that the information could only be viewed by the searcher, and that Cuil would not let other search engines access the Facebook information, according to Seval Oz Ozveren, a Cuil vice president. Facebook Results is the first such release between Cuil and a social networking site to integrate users' social profile on search pages. It was announced in November; the concept was first discussed by Cuil in July. More such deals are expected to follow, she says.

"Social search is here to stay, and we are certain to see more Facebook integration by other players as well," says Oren Etzioni, a computer scientist and search researcher at the University of Washington, who added that Facebook's permissions will be the key to such efforts. "We see how important Facebook and other social networks are, and we also see how Facebook is seeking to parlay that importance into a role on other sites using initiatives like Facebook Connect, and now this one."