Advancements in artificial intelligence, particularly in computer vision, have driven the research and development of visual food analysis systems focused primarily on enhancing people’s well-being. Food analysis can be performed at various levels of granularity, with food segmentation being a major component of numerous real-world applications. Deep learning-based methodologies have demonstrated promising results in food segmentation; however, many of these approaches demand high computational resources, making them impractical for low-performance devices. In this research, a novel, lightweight, deep learning-based method for semantic food segmentation is proposed. To achieve this, the state-of-the-art DeepLabv3+ model was adapted by optimizing the backbone with the lightweight network EfficientNet-B1, replacing the Atrous Spatial Pyramid Pooling (ASPP) in the neck with Cascade Waterfall ASPP (CWASPP), and refining the encoder output using the squeeze-and-excitation attention mechanism. To validate the method, four publicly available food datasets were selected. Additionally, a new food segmentation dataset consisting of self-acquired food images was introduced and included in the validation. The results demonstrate that high performance can be achieved at a significantly lower cost. The proposed method yields results that are either better than or comparable to those of state-of-the-art techniques while requiring significantly lower computational costs. In conclusion, this research demonstrates the potential of deep learning to perform food image segmentation on low-performance stand-alone devices, paving the way for more efficient, cost-effective, and scalable food analysis applications.
6 September 2022 in IEEE Latin America Transactions
The emergence of deep learning models has made it possible to address several real-life problems and, in particular, those in which computer vision plays a key role. In this sense, the food recognition task from images is one of the beneficiaries of this machine learning method. Its importance lies in its usefulness to become aware of the food eaten and in this way help us to lead a healthy lifestyle. In recent years, food image recognition has gained great prominence in the literature, providing novel models and datasets to address it. However, public data generally correspond to American, Asian, and European foods, therefore the methods developed cannot be directly applied to the Chilean diet. In this article we will publish a new dataset to recognize the foods present in the Chilean diet. In addition, we will perform a comparison with public popular food datasets to analyze the similarity in the dishes of the proposed dataset with respect to the exciting ones in the literature. Moreover, we will establish a baseline using the state-of-the-arts Convolutional Neural Network architectures and the novel Swin Transformer approach.