A joint end-to-end learning method for multi-view object instance detection and re-identification that incorporates learned geometric soft constraints, validated on a new street-level panorama dataset.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Simultaneous multi-view instance detection with learned geometric soft-constraints
A joint end-to-end learning method for multi-view object instance detection and re-identification that incorporates learned geometric soft constraints, validated on a new street-level panorama dataset.