Dietary studies showed that dietary problems such as obesity are associated with other chronic diseases, including hypertension, irregular blood sugar levels, and increased risk of heart attacks. The primary cause of these problems is poor lifestyle choices and unhealthy dietary habits, which are manageable using interactive mHealth apps. However, traditional dietary monitoring systems using manual food logging suffer from imprecision, underreporting, time consumption, and low adherence. Recent dietary monitoring systems tackle these challenges by automatic assessment of dietary intake through machine learning methods. This survey discusses the best-performing methodologies that have been developed so far for automatic food recognition and volume estimation. Firstly, the paper presented the rationale of visual-based methods for food recognition. Then, the core of the study is the presentation, discussion, and evaluation of these methods based on popular food image databases. In this context, this study discusses the mobile applications that are implementing these methods for automatic food logging. Our findings indicate that around 66.7% of surveyed studies use visual features from deep neural networks for food recognition. Similarly, all surveyed studies employed a variant of convolutional neural networks (CNN) for ingredient recognition due to recent research interest. Finally, this survey ends with a discussion of potential applications of food image analysis, existing research gaps, and open issues of this research area. Learning from unlabeled image datasets in an unsupervised manner, catastrophic forgetting during continual learning, and improving model transparency using explainable AI are potential areas of interest for future studies.

State-of-the-art deep learning models for food recognition do not allow data incremental learning and often suffer from catastrophic interference problems during the class incremental learning. This is an important issue in food recognition since real-world food datasets are open-ended and dynamic, involving a continuous increase in food samples and food classes. Model retraining is often carried out to cope with the dynamic nature of the data, but this demands high-end computational resources and significant time. This paper proposes a new open-ended continual learning framework by employing transfer learning on deep models for feature extraction, Relief F for feature selection, and a novel adaptive reduced class incremental kernel extreme learning machine (ARCIKELM) for classification. Transfer learning is beneficial due to the high generalization ability of deep learning features. Relief F reduces computational complexity by ranking and selecting the extracted features. The novel ARCIKELM classifier dynamically adjusts network architecture to reduce catastrophic forgetting. It addresses domain adaptation problems when new samples of the existing class arrive. To conduct comprehensive experiments, we evaluated the model against four standard food benchmarks and a recently collected Pakistani food dataset. Experimental results show that the proposed framework learns new classes incrementally with less catastrophic inference and adapts domain changes while having competitive classification performance.

Background: Diet monitoring has been linked with improved eating habits and positive health outcomes such as prevention of obesity. However, this is often unsustainable as traditional methods place a high burden on both participants and researchers through pen and paper recordings and manual nutrient coding respectively. The digitisation of dietary monitoring has greatly reduced these barriers. This paper proposes a diet application with a novel food recognition feature with a usability study conducted in the real world. Methods: This study describes the development of a mobile diet application (MyDietCam) targeted at healthy Malaysian adults. Focus group discussions (FGD) were carried out among dietitians and potential users to determine ideal features in a diet application. Thirty participants were recruited from a local university to log their meals through MyDietCam for six days and submit the Malay mHealth Application Usability Questionnaire (M-MAUQ) at the end of the study. Results: The findings from the FGD led to the implementation of the main features: individualised recommendations, food logging through food recognition to reduce steps for data entry and provide detailed nutrient analyses through visuals. An average overall usability score of 5.13 out of a maximum of seven was reported from the M-MAUQ which is considered acceptable. Conclusion: The development of a local (Malaysian) mobile diet application with acceptable usability may be helpful in sustaining the diet monitoring habit to improve health outcomes. Future work should focus on improving the issues raised before testing the effectiveness of the application for improving health outcomes.

Recently, food recognition has received more research attention for mHealth applications that use automated visual-based methods to assess dietary intake. The goal is to improve the food diaries by addressing the challenges faced by existing methodologies. In addition to the classical challenge of the absence of rigid food structure and intra-class variations, food diaries employing deep networks trained with pristine images are susceptible to quality variations in real-world conditions of image acquisition and transmission. Similarly, existing progressive classifiers that use visual features via a convolutional neural network (CNN) classify food categories and cannot detect food ingredients. We aim to provide a system that selects the optimal subset of features from quality resilient CNNs and subsequently incorporates the parallel type of classification to tackle such challenges. The first progressive classifier recognizes food categories, and its multilabel extension detects food ingredients. Following this idea, after extracting features from the quality resilient category and ingredient CNN models by fine-tuning it on synthetic images generated using the novel online data augmentation method random iterative mixup. Our feature selection strategy uses the Shapley additive explanation (SHAP) values from the gradient explainer to select the best features. Then, novel progressive kernel extreme learning machine (PKELM) is exploited to cater to domain variations due to quality distortions, intra-class variations, and so forth, by remodeling the network structure based on activity value with the nodes. PKELM extension for multilabel classification detects ingredients by employing a bipolar step function to process test output and then selecting the column labels of the resulting matrix with a value of one. Moreover, during online learning, the PKELM novelty detection mechanism can label unlabeled instances and detect noisy samples. Experimental results showed superior performance on an integrated set of measures for seven publicly available food datasets.

With the exponential growth of the Internet population, scientists and researchers face the large-scale data for processing. However, the traditional algorithms, due to their complex computation, are not suitable for the large-scale data, although they play a vital role in dealing with large-scale data for classification and regression. One of these variants, which is called Reduced Kernel Extreme Learning Machine (Reduced-KELM), is widely used in the classification task and attracts attention from researchers due to its superior performance. However, it still has limitations, such as instability of prediction because of the random selection and the redundant training samples and features because of large-scaled input data. This study proposes a novel model called Reformed Reduced Kernel Extreme Learning Machine with RELIEF-F (R-RKELM) for human activity recognition. RELIEF-F is applied to discard the attributes of samples with the negative values in the weights. A new sample selection approach, which is used to further reduce training samples and to replace the random selection part of Reduced-KELM, solves the unstable classification problem in the conventional Reduced-KELM and computation complexity problem. According to experimental results and statistical analysis, our proposed model obtains the best classification performances for human activity data sets than those of the baseline model, with an accuracy of 92.87 % for HAPT, 92.81 % for HARUS, and 86.92 % for Smartphone, respectively.