This project involves developing a deep learning model to accurately recognize food items from images and estimate their calorie content. The primary goal is to enable users to track their dietary intake and make informed food choices.
The project uses the Food-101 dataset, which consists of 101 different food categories with a total of 101,000 images. The dataset is utilized to train a convolutional neural network (CNN) for food item classification. Additionally, a separate dataset (food.csv) is used to estimate the calorie content of the recognized food items based on the food names.
- Food Item Recognition: Classify images of food into one of 101 categories using a convolutional neural network.
- Calorie Estimation: Predict the calorie content of recognized food items from a provided dataset.
- Evaluation: Achieved an accuracy of 52% in classifying food items.
- Example Predictions: Demonstrates how the model can accurately recognize food items and estimate their calorie content.