Register now After registration you will be able to apply for this opportunity online.
Master Thesis internship with WayBetter Inc. in collaboration with ETH Zurich (6-months duration)
This six-month internship at WayBetter Inc., in collaboration with ETH Zurich, involves a cutting-edge machine learning project to develop an AI model that detects weight changes through facial images using a unique dataset of 6 million labeled full-body images. This model aims to facilitate significant applications in telehealth and clinical monitoring. Candidates will have the option to integrate this project into their Master's thesis at ETH Zurich, benefitting from expert guidance while contributing to transformative health monitoring solutions. Ideal candidates should have a solid foundation in machine learning, image processing, and data management.
Keywords: Machine Learning, Computer Vision, Digital Health
WayBetter, a leader in technology-driven weight management solutions, is seeking a highly motivated master’s student to embark on a cutting-edge machine learning project. This project aims to harness WayBetter’s expansive dataset of labeled full-body images to develop an AI model capable of detecting weight changes through facial images. This initiative holds significant clinical and telehealth applications, such as monitoring weight changes in elderly patients remotely or alerting doctors during telehealth sessions about significant weight fluctuations.
**Dataset**:
The dataset includes approximately 6 million full-body images, each annotated with weight, height, and sex labels, verified by human oversight. This unique collection spans over 10 years, capturing the weight-loss journeys of numerous individuals. Each subject in the dataset has multiple images, showing them at different stages of their weight loss, providing a rich source for training models on subtle physical changes. The images are of high quality, taken in consistent lighting conditions, and with the subjects in lightweight clothing, ensuring clarity and consistency.
**Problem Statement**:
Previous studies have attempted to estimate body weight from images, but none have had access to a dataset as extensive and detailed as WayBetter’s. This project seeks to leverage this unparalleled dataset to train a model that can not only estimate weight changes from facial images but do so with a focus on detecting clinically meaningful changes. The goal is not to predict exact weight but to identify significant weight changes that could indicate health issues or the effectiveness of a health regimen.
**Project Aims**:
- Model Development: Develop and train machine learning models to accurately detect weight changes from facial images.
- Accuracy and Validation: Evaluate the model's accuracy in detecting significant weight changes, using a subset of the dataset for validation.
- Clinical Application: Assess the model’s utility in practical scenarios, such as remote patient monitoring and telehealth.
- Scalability: Explore the scalability of the solution for broader application beyond the initial clinical contexts.
**Candidate Profile**:
Candidates for this project should have a strong foundation in machine learning and
data modeling, with specific skills and experience in:
- Deep Learning: Proficiency in CNNs (Convolutional Neural Networks) and potentially other neural network architectures.
- Image Processing: Experience with image analysis, facial recognition technologies, or related fields.
- Data Handling: Ability to manage and process large datasets efficiently.
- Programming: Strong coding skills in Python, including familiarity with ML libraries like TensorFlow or PyTorch.
- Statistical Analysis: Competence in applying statistical methods to analyze and interpret complex datasets.
- Creativity and Innovation: Ability to think creatively about new approaches to challenging problems.
**Conclusion**:
This project represents a unique opportunity to work on a high-impact, real-world
problem using state-of-the-art machine learning techniques and a comprehensive dataset. The successful candidate will not only work on advanced AI techniques but will also contribute to potentially transformative healthcare solutions. Interested students are encouraged to apply, bringing their expertise and enthusiasm to tackle this exciting challenge. Together with WayBetter, you can help shape the future of health monitoring and intervention through innovative AI-driven solutions.
**Master Thesis Collaboration**:
In collaboration with the Centre for Digital Health Interventions (C4DHI) at ETH Zurich, applicants are encouraged to integrate their internship project into their master's thesis. Throughout their project, they will receive guidance from two doctoral researchers at ETH, enriching their academic journey with practical, real-world experience.
**Application**:
Applicants can apply via e-mail to rjakob@ethz.ch, also including their CV and a motivation letter highlighting relevant past experiences regarding the position. Students from universities other than ETH are also encouraged to apply.
WayBetter, a leader in technology-driven weight management solutions, is seeking a highly motivated master’s student to embark on a cutting-edge machine learning project. This project aims to harness WayBetter’s expansive dataset of labeled full-body images to develop an AI model capable of detecting weight changes through facial images. This initiative holds significant clinical and telehealth applications, such as monitoring weight changes in elderly patients remotely or alerting doctors during telehealth sessions about significant weight fluctuations.
**Dataset**: The dataset includes approximately 6 million full-body images, each annotated with weight, height, and sex labels, verified by human oversight. This unique collection spans over 10 years, capturing the weight-loss journeys of numerous individuals. Each subject in the dataset has multiple images, showing them at different stages of their weight loss, providing a rich source for training models on subtle physical changes. The images are of high quality, taken in consistent lighting conditions, and with the subjects in lightweight clothing, ensuring clarity and consistency.
**Problem Statement**: Previous studies have attempted to estimate body weight from images, but none have had access to a dataset as extensive and detailed as WayBetter’s. This project seeks to leverage this unparalleled dataset to train a model that can not only estimate weight changes from facial images but do so with a focus on detecting clinically meaningful changes. The goal is not to predict exact weight but to identify significant weight changes that could indicate health issues or the effectiveness of a health regimen.
**Project Aims**:
- Model Development: Develop and train machine learning models to accurately detect weight changes from facial images. - Accuracy and Validation: Evaluate the model's accuracy in detecting significant weight changes, using a subset of the dataset for validation. - Clinical Application: Assess the model’s utility in practical scenarios, such as remote patient monitoring and telehealth. - Scalability: Explore the scalability of the solution for broader application beyond the initial clinical contexts.
**Candidate Profile**: Candidates for this project should have a strong foundation in machine learning and data modeling, with specific skills and experience in:
- Deep Learning: Proficiency in CNNs (Convolutional Neural Networks) and potentially other neural network architectures. - Image Processing: Experience with image analysis, facial recognition technologies, or related fields. - Data Handling: Ability to manage and process large datasets efficiently. - Programming: Strong coding skills in Python, including familiarity with ML libraries like TensorFlow or PyTorch. - Statistical Analysis: Competence in applying statistical methods to analyze and interpret complex datasets. - Creativity and Innovation: Ability to think creatively about new approaches to challenging problems.
**Conclusion**: This project represents a unique opportunity to work on a high-impact, real-world problem using state-of-the-art machine learning techniques and a comprehensive dataset. The successful candidate will not only work on advanced AI techniques but will also contribute to potentially transformative healthcare solutions. Interested students are encouraged to apply, bringing their expertise and enthusiasm to tackle this exciting challenge. Together with WayBetter, you can help shape the future of health monitoring and intervention through innovative AI-driven solutions.
**Master Thesis Collaboration**: In collaboration with the Centre for Digital Health Interventions (C4DHI) at ETH Zurich, applicants are encouraged to integrate their internship project into their master's thesis. Throughout their project, they will receive guidance from two doctoral researchers at ETH, enriching their academic journey with practical, real-world experience.
**Application**: Applicants can apply via e-mail to rjakob@ethz.ch, also including their CV and a motivation letter highlighting relevant past experiences regarding the position. Students from universities other than ETH are also encouraged to apply.
This project seeks to leverage an existing longitudinal dataset to train a model that can estimate weight changes from facial images. The focus is on detecting clinically meaningful changes, and the goal is to identify significant weight changes that could indicate health issues or the effectiveness of a health regimen.
This project seeks to leverage an existing longitudinal dataset to train a model that can estimate weight changes from facial images. The focus is on detecting clinically meaningful changes, and the goal is to identify significant weight changes that could indicate health issues or the effectiveness of a health regimen.