People who find it difficult to lose weight may benefit from a voice-controlled calorie counter, scientists say.

Obesity remains a concern for health experts as numbers of obese and overweight individuals continue to rise steadily. An earlier study revealed that doctor's advice about eating right and exercising daily is not enough to lower obesity rates across the population.

Logging daily food intake helps in checking whether an individual is meeting or exceeding their caloric intake; however, this entails the accuracy of the caloric count and the consistency of logging to be effective in weight loss. Many people often forget logging in their intake or second-guess caloric count and portions, leading to failure in weight loss.

Tufts University nutritionists were experimenting with a smartphone app that can be used for caloric intake recording when they thought of collaborating with the Massachusetts Institute of Technology's (MIT) Spoken Language System Group of Computer Science and Artificial Intelligence Laboratory (CSAIL) to create a voice-controlled application that aims to make daily intake logging more user-friendly.

The application would allow the user to describe a particular meal and the speech-controlled nutrition logging system would use data from the U.S. Department of Agriculture (USDA) to retrieve nutritional data.

As data is pulled from the registry, the food will be displayed together with the corresponding menus that will allow the user to refine the food description, such as portions and descriptions. Refinements, according to the scientists, can also be through a voice command.

Senior research scientist James Glass of the CSAIL said that existing apps in the market only allow people to log meals in such a tedious manner that consistency becomes an issue, thus their development of the application that aims to help users accurate food information easily.

Researchers are fine-tuning the application to properly recognize the context of the food being logged. The system must be able to differentiate the functional role of the verbal cues of the user. For example, if the user says "bowl of oatmeal," it should be properly differentiated from "oatmeal cookie" in order to pull out the accurate information from the registry. Researchers used machine learning algorithms to look for syntactic relationship patterns between words and their functional roles to address the said problem.

Another issue that researchers are trying to resolve is the "user phrasing" with that of the USDA database. To resolve this problem, a semantic matching was developed through the help of an open-source database, Freebase, that has a list of more than 8,000 common food items, including their synonyms.

Experts believe that food logging is essential in keeping the pounds off. Susan Roberts of Tufts' USDA-sponsored Jean Mayer Human Nutrition Research Center on Aging said that food logging increases awareness of food intake, which in turn gives people an idea of how many calories a small serving can have. She acknowledged that existing apps are helpful but can be time-consuming to use, resulting in inconsistency and ineffective weight loss.

"A spoken language system that you can use with your phone would allow people to log food wherever they are eating it, with less work. As I see it, we need to come up with something that really isn't much work, so it isn't an extra burden in life," Roberts added.

A Web-based prototype of the said collaboration was presented at the International Conference on Acoustics, Speech, and Signal Processing in Shanghai by the MIT researchers.

The Centers for Disease Control and Prevention states that even modest weight loss can significantly improve the health of previously obese and overweight individuals. A weight loss of about 5-10 percent of the total body weight can improve blood cholesterol levels, maintain blood pressure and lower blood sugar.

Photo: Julia Lamphear | Flickr

ⓒ 2021 All rights reserved. Do not reproduce without permission.