What a pregnant lady should eat: a machine learning empowered web-based application to detect food

Abstract

It's critical for the growth of the fetus as well as the health of the mother to get enough nutrients throughout pregnancy. Women need special attention throughout pregnancy. Maintaining a balanced diet and sufficient nutrient intake are essential. Unfortunately, a significant percentage of Bangladeshis are unaware of this issue. The people that suffer the most are new moms, as they have never encountered this circumstance before. Major pregnancy issues might result from an unbalanced diet and incorrect fruit consumption at this period. To overcome this challenge, we applied a deep learning technique called You Only Look Once (YOLOv8) to identify and categorize various fruit varieties. A varied dataset of foods pertinent to prenatal nutrition is used to modify and train the YOLOv8 model, which is well-known for its real-time object identification skills. The system's goal is to make it easier for expectant mothers to keep an eye on their food consumption and encourage a nutritious, well-balanced diet for the duration of their pregnancy. We have separated our eight fruit classes into two main groups according to whether they are avoided or useful. We obtained data about fruit eating during pregnancy from online health journals and papers as well as gynecologists. Reading our article will help people understand which of our eight fruit types to consume and which to avoid. Pre-trained YOLOV8 models were implemented. This study contributes to the field of maternal healthcare by providing expectant mothers with a novel tool that facilitates informed dietary decisions. The YOLOv8-based food detection technology enhances maternal and fetal health outcomes while also increasing knowledge of nutrition. Further research endeavors might concentrate on enhancing the efficacy of the model, integrating customized nutritional suggestions, and investigating its utilization in more comprehensive healthcare settings.