Retrieval-augmented generation (RAG) for computer vision applications
room
TNO
|
Den Haag, Netherlands, Europe

Explore the potential of RAG for computer vision, e.g. image classification. Research the methodologies to incorporate RAG into computer vision models. Investigate how RAG can be effectively leveraged in smaller datasets. Explore for which sizes of datasets and various degrees of class imbalance RAG can be helpful. Explore which other tasks RAG could be relevant for (e.g. one/few-shot learning, continual learning).

share
Share