Researchers from Penn State have identified a link between fast eating rates in children and increased obesity risk. Traditional methods for measuring bite rates are labor-intensive and limit data collection. To address this, the Penn State team developed an AI model, named ByteTrack, capable of automatically counting bites during meals. A recent pilot study published in Frontiers in Nutrition showed the AI’s bite count accuracy is about 70% compared to human observers, with further improvements needed for real-world application.
Faster eating may prevent children from feeling full, promoting overeating. The AI model aims to assist researchers, parents, and health professionals in identifying when children should slow down their eating. Using video footage of children eating, researchers trained ByteTrack to detect bites while recognizing faces.
Future applications may include a smartphone app to help children develop healthier eating habits, ultimately aiming to combat childhood obesity and promote long-term wellness.
