Fraunhofer

Masterarbeit: Modellierungsansätze und Loss-Design für präzise Altersschätzung

Fraunhofer

Darmstadt · Posted Apr 23

Other Onsite
Apply now

Master thesis position at Fraunhofer SIT in Darmstadt, Germany, focusing on modeling approaches and loss design for precise age estimation from facial images. Systematic comparison of regression (point, probabilistic, quantile), classification, ordinal regression, label distribution learning, and hybrid methods; investigation of missing/small age classes, imbalance, label noise, and non-linear age scale; development and evaluation of robust methods; aims for reproducible implementations and strong baselines with evaluation on UTKFace, IMDB-WIKI, APPA-REAL, MORPH II, AgeDB.

What they are looking for

Python Pytorch Opencv Computer vision Machine learning

Details

Work type
Onsite
Remote eligible
No

Get new internships by email

Free daily digest, matched to what you pick. Unsubscribe anytime.