Die KI-Suchmaschine für Jobs
Master's Thesis: Fair and Balanced Age Estimation through Dynamic Group Training(m/w/x)
Developing and evaluating dynamic training strategies for fair age estimation, implementing oversampling and curriculum strategies at application-oriented research organization. Good machine learning and neural network training knowledge required, PyTorch/OpenCV experience preferred. Independent schedule management, insights into research and industry.
Anforderungen
- Good knowledge in machine learning and neural network training
- Ideally, knowledge in computer vision and facial recognition
- Good Python skills, preferably experience with PyTorch, OpenCV
- Motivation for independent research into new topics
- Interest in robustness and evaluation metrics
- Interest in scientific research
Aufgaben
- Develop and systematically evaluate dynamic training strategies
- Calculate and use subgroup-specific metrics for control
- Develop and implement dynamic oversampling strategies
- Develop and implement uncertainty sampling strategies
- Develop and implement curriculum strategies
- Examine appropriate aggregation metrics over subgroups
- Compare with classic oversampling, probabilistic sampling, GroupDRO, and JTT
- Explore combinations using AutoML and hyperparameter search
- Evaluate methods using freely available benchmark datasets
- Compare developed methods with existing approaches
- Document code for reusability and reproducibility
- Research and compile information on current ML topics
- Research and implement novel ML and computer vision approaches
- Self-critically evaluate obtained results
- Present the results
- Prepare a scientific paper as a master's thesis
Ausbildung
- Laufendes StudiumODER
- Bachelor-AbschlussODER
- Master-Abschluss
Sprachen
- Englisch – verhandlungssicher
Tools & Technologien
- Python
- PyTorch
- OpenCV
Benefits
Flexibles Arbeiten
- Independent work schedule management
Startup-Atmosphäre
- Insights into research and industry
Noch nicht perfekt?
- Fraunhofer-GesellschaftVollzeitPraktikumnur vor OrtDarmstadt
- Fraunhofer-Gesellschaft
Master's Thesis: Modelling Approaches and Loss Design for Precise Age Estimation(m/w/x)
Vollzeitnur vor OrtKeine AngabeDarmstadt - Fraunhofer-Gesellschaft
Masterarbeit: Modellierungsansätze und Loss-Design für präzise Altersschätzung(m/w/x)
VollzeitPraktikumnur vor OrtDarmstadt - Fraunhofer-Gesellschaft
Master's Thesis: Detection of Suggestive Poses and Facial Expressions(m/w/x)
VollzeitPraktikumnur vor OrtDarmstadt - Fraunhofer-Gesellschaft
Master's Thesis: Detection and Segmentation of Suggestive Clothing(m/w/x)
VollzeitPraktikumnur vor OrtDarmstadt
Master's Thesis: Fair and Balanced Age Estimation through Dynamic Group Training(m/w/x)
Developing and evaluating dynamic training strategies for fair age estimation, implementing oversampling and curriculum strategies at application-oriented research organization. Good machine learning and neural network training knowledge required, PyTorch/OpenCV experience preferred. Independent schedule management, insights into research and industry.
Anforderungen
- Good knowledge in machine learning and neural network training
- Ideally, knowledge in computer vision and facial recognition
- Good Python skills, preferably experience with PyTorch, OpenCV
- Motivation for independent research into new topics
- Interest in robustness and evaluation metrics
- Interest in scientific research
Aufgaben
- Develop and systematically evaluate dynamic training strategies
- Calculate and use subgroup-specific metrics for control
- Develop and implement dynamic oversampling strategies
- Develop and implement uncertainty sampling strategies
- Develop and implement curriculum strategies
- Examine appropriate aggregation metrics over subgroups
- Compare with classic oversampling, probabilistic sampling, GroupDRO, and JTT
- Explore combinations using AutoML and hyperparameter search
- Evaluate methods using freely available benchmark datasets
- Compare developed methods with existing approaches
- Document code for reusability and reproducibility
- Research and compile information on current ML topics
- Research and implement novel ML and computer vision approaches
- Self-critically evaluate obtained results
- Present the results
- Prepare a scientific paper as a master's thesis
Ausbildung
- Laufendes StudiumODER
- Bachelor-AbschlussODER
- Master-Abschluss
Sprachen
- Englisch – verhandlungssicher
Tools & Technologien
- Python
- PyTorch
- OpenCV
Benefits
Flexibles Arbeiten
- Independent work schedule management
Startup-Atmosphäre
- Insights into research and industry
Über das Unternehmen
Fraunhofer-Gesellschaft
Branche
Research
Beschreibung
Das Unternehmen ist eine der führenden Organisationen für anwendungsorientierte Forschung mit 76 Instituten in Deutschland.
Noch nicht perfekt?
- Fraunhofer-Gesellschaft
Masterarbeit: Faire und ausbalancierte Altersschätzung durch dynamisches, gruppenweises Training(m/w/x)
VollzeitPraktikumnur vor OrtDarmstadt - Fraunhofer-Gesellschaft
Master's Thesis: Modelling Approaches and Loss Design for Precise Age Estimation(m/w/x)
Vollzeitnur vor OrtKeine AngabeDarmstadt - Fraunhofer-Gesellschaft
Masterarbeit: Modellierungsansätze und Loss-Design für präzise Altersschätzung(m/w/x)
VollzeitPraktikumnur vor OrtDarmstadt - Fraunhofer-Gesellschaft
Master's Thesis: Detection of Suggestive Poses and Facial Expressions(m/w/x)
VollzeitPraktikumnur vor OrtDarmstadt - Fraunhofer-Gesellschaft
Master's Thesis: Detection and Segmentation of Suggestive Clothing(m/w/x)
VollzeitPraktikumnur vor OrtDarmstadt