If you are interested in working on one of the projects listed below, please send an email to Prof. Marcel Salathe in which you answer the following questions:
- Which project(s) are you interested in?
- Are you looking for a thesis or a semester project / Lab immersion? Priority will be given to Master Thesis Projects (full-time)
- Why did these specific projects catch your attention?
- How much time could you spend on the project per week? How many classes are you planning on taking in addition to the project?
Please attach your CV and links to projects you developed (e.g. GitHub repo)
Background:
The accurate measurement of food volume is critically important in nutritional science and food technology. It’s essential for tasks like dietary assessment, portion size measurement, and monitoring caloric intake. Traditional volume estimation methods, which often involve manual measurements or basic assumptions, may not accurately represent the diverse shapes and sizes of various food items.
Project Description:
This project is designed to utilize deep learning for estimating the volume of food items from images. The project utilizes a dataset comprising 100,000 images of different food items. Each image is meticulously annotated, detailing the segmentation of each food item, classification of these segments, and the estimated weight of each item. The main goal is to develop a model that precisely estimates food volume, using this rich dataset and advanced deep learning techniques.
Model Development:
Explore and apply state-of-the-art deep learning models, while considering using multi-modal inputs, combining image data with meta-data (classification labels and weights) for more accurate volume estimation.
Volume Estimation Strategy:
Formulate a method to calculate the volume of food items, either directly from segmentation masks or indirectly by correlating known weights with estimated volumes.
Use geometric and spatial analysis to refine volume estimation, taking into account the shape and orientation of food items in the images.
Model Training and Evaluation:
Train the model with suitable loss functions and optimization methods, focusing on accurate volume estimation.nAssess the model’s accuracy by comparing predicted volumes with manually measured volumes for a portion of the dataset.
Results Analysis and Optimization :
Evaluate the results to determine the model’s strengths and weaknesses.Improve the model architecture, training procedure, and volume estimation approach based on this evaluation, aiming to increase the precision of the volume estimations.
Expected Outcomes:
The successful completion of this project is anticipated to produce a robust deep learning model that can precisely estimate the volume of various food items from images. This advancement promises to significantly enhance areas such as nutritional science, dietary management, and food technology, offering a scalable and automated approach to food volume estimation. Moreover, the insights from this project may open doors to further research and applications in related fields like automated dietary tracking and smart food preparation systems.
Project Background:
Continuous Glucose Monitoring (CGM) captures detailed information about glucose levels over time. Our Food and You dataset contains CGM data from nearly 1000 individuals, with at least two weeks of CGM readings. From this data, various glucose metrics can be derived using the iGLU package in R, such as mean glucose amplitude, glycemic variability, high/low blood glucose index etc., which provides insights into glucose variability and stability. Understanding how glucose metrics stabilize over time, the effects of weekends on glucose control, and the impact of demographic factors such as age, gender, and BMI and dietary factors is critical to personalizing and improving glucose management strategies.
Project Aim:
This project aims to explore several aspects of glucose variability in relation to time, demographics, and lifestyle, focusing on three primary questions:
1. Minimum Data Requirements: How many days of CGM data are necessary for the reliable estimation of various glucose metrics? Do the collected need to be continuous or discontinuous for estimating reliability?
2. Temporal Patterns: Are there significant changes in glucose behavior over weekends or specific time periods? Frequency of hypoglycemic and hyperglycemic events?
3. Demographic Differences: How do factors such as age, gender, and nutritional intake influence glucose control and variability?
Expected Outcomes
– Identification of the minimum number of days needed to capture stable glucose metrics.
– Insights into temporal patterns, particularly weekend effects, on glucose control.
– A deeper understanding of how demographic factors affect glucose variability and responses.
Skills Acquired:
– Data analysis using CGM data.
– Experience with statistical techniques such as mixed-effects models, time-series analysis, and hypothesis testing.
– Insight into metabolic health and personalized medicine.
Requirement:
– Familiarity with Python/R is recommended.
– Basic understanding of time-series data and metabolic health would be beneficial.
Project Duration
6 months
Project Background:
Physical activity (PA) is a crucial component of maintaining a healthy lifestyle, and understanding the patterns and accuracy of reported activity levels can provide valuable insights for health interventions. In our Food and You dataset, which collected nutritional data for at least two weeks, we also collected user-reported (subjective) PA and smart-watch (objective) PA data for nearly one-third of the participants. These complementary data sources offer a unique opportunity to assess activity levels, validate subjective reports, investigate the influences of demographic factors on physical behavior, and explore associations between physical activity and dietary habits.
Project Aim:
This project aims to do data processing and investigating multiple aspects of physical activity using both objective and subjective data, focusing on the following questions:
1. Demographic Comparisons: How do physical activity levels and types of activities vary among different demographic groups (e.g., age, gender, BMI)?
2. Behavioral Insights: What are the temporal patterns of PA in participants? i.e., frequency of PA in a week, ratio of high intensity vs low intensity days etc. Are there major differences between self-reported and objectively measured activity?
3. Diet and Activity Associations: How are different types and levels of PA associated with dietary patterns captured in the dataset? Are some demographics more likely to overestimate or underestimate their physical activity?
Skills Acquired:
– Experience in data analysis involving both objective and subjective physical activity measures.
– Application of statistical techniques to compare and validate different data types (e.g., correlation analysis, ICC, mixed effects models, etc).
– Graph visualization to effectively communicate insights.
Requirements:
– Familiarity with statistical analysis in Python/R is recommended.
– Basic knowledge of public health concepts and interest in behavioral analysis.
Project Duration:
6 months