Nutrition and Computer Vision
Motivation and scope
Nutrition has a huge but not yet fully explored impact on human health. To fully measure and assess the impact of nutrition on human health, it is necessary to measure food intake accurately. The existing methodologies are limited by tedious and error-prone data collection process. However, recent developments in computer vision, depth imagining and deep learning, can substantially improve this process. The objective of this thesis is to compare the existing methods for nutrition evaluation and assess their applicability for food type recognition and volume estimation.
The critical task is to compare the ability of the state of the art algorithms to differentiate different types of food and estimate their mass/volume using RGB/RGB-D images.
Good programming skills in Python/C++. Basic knowledge in machine learning and computer vision.
You will learn how to work on a challenging machine learning project using deep learning algorithms for image classification and volume estimation.