In this talk we will revisit earlier work on people/object tracking and show how it can be framed into a coherent picture within a modern deep learning approach. We will consider invariance and equivariance as mathematical principles to derive our recurrent deep network architecture for multi-object tracking.
Learning to Track (extended abstract of MCV18 invited talk)
Lanz, Oswald
2018-01-01
Abstract
In this talk we will revisit earlier work on people/object tracking and show how it can be framed into a coherent picture within a modern deep learning approach. We will consider invariance and equivariance as mathematical principles to derive our recurrent deep network architecture for multi-object tracking.File in questo prodotto:
Non ci sono file associati a questo prodotto.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
