FoodNet: Recognizing Foods Using Ensemble of Deep Networks

TL;DR

FoodNet introduces an ensemble deep learning approach for automatic food recognition, combining features from multiple CNNs and classical methods, achieving high accuracy on large food image datasets.

Contribution

The paper presents a novel ensemble CNN architecture that leverages multiple deep networks and handcrafted features for improved food classification accuracy.

Findings

Achieved superior accuracy on ETH Food-101 dataset.

Demonstrated effectiveness on a new Indian food image database.

Outperformed several benchmark CNN frameworks.

Abstract

In this work we propose a methodology for an automatic food classification system which recognizes the contents of the meal from the images of the food. We developed a multi-layered deep convolutional neural network (CNN) architecture that takes advantages of the features from other deep networks and improves the efficiency. Numerous classical handcrafted features and approaches are explored, among which CNNs are chosen as the best performing features. Networks are trained and fine-tuned using preprocessed images and the filter outputs are fused to achieve higher accuracy. Experimental results on the largest real-world food recognition database ETH Food-101 and newly contributed Indian food image database demonstrate the effectiveness of the proposed methodology as compared to many other benchmark deep learned CNN frameworks.