Penn州人工智能系统跟踪儿童被咬率,通过鼓励吃慢,帮助降低肥胖风险。 A Penn State AI system tracks kids' bite rates to help reduce obesity risk by encouraging slower eating.

宾夕法尼亚州立大学研究人员开发的名为 " ByteTrack " 的新的人工智能系统旨在通过跟踪膳食时的咬牙率来消除儿童肥胖症。 A new AI system called ByteTrack, developed by Pennsylvania State University researchers, aims to combat childhood obesity by tracking bite rates during meals. 该人工智能根据 94 名 7 至 9 岁儿童的 1,440 分钟视频进行了训练，识别人脸的准确率为 97%，但大约 70% 的时间正确计算咬伤次数，与被遮挡的脸或咀嚼餐具的儿童作斗争。 Trained on 1,440 minutes of video from 94 children aged 7 to 9, the AI identifies faces with 97% accuracy but counts bites correctly about 70% of the time, struggling with obscured faces or children chewing utensils. 较快进食与由于全面性信号降低而造成的肥胖风险较高有关。 Faster eating is linked to higher obesity risk due to reduced fullness signals. 该系统最终可以给智能电话应用程序提供动力,促使儿童吃得较慢,支持健康习惯。 The system could eventually power smartphone apps to prompt children to eat slower, supporting healthier habits. 该研究发表在《营养前沿》上,专家说,有针对性地点食率可能有助于减少过量饮食和肥胖风险。 The study was published in Frontiers in Nutrition, and experts say targeting bite rate may help reduce overeating and obesity risk.