We present a unified framework for studying the identifiability of representations learned from simultaneously observed views, such as different data modalities.
We present a unified framework for studying the identifiability of representations learned from simultaneously observed views, such as different data modalities.