Timezone: »

 
Contributed Talk 2: Witness Autoencoder: Shaping the Latent Space with Witness Complexes
Anastasiia Varava · Danica Kragic · Simon Schönenberger · Jen Jen Chung · Roland Siegwart · Vladislav Polianskii

Fri Dec 11 07:06 AM -- 07:11 AM (PST) @
in Differential Geometry meets Deep Learning (DiffGeo4DL) »

We present a Witness Autoencoder (W-AE) – an autoencoder that captures geodesic distances of the data in the latent space. Our algorithm uses witness complexes to compute geodesic distance approximations on a mini-batch level, and leverages topological information from the entire dataset while performing batch-wise approximations. This way, our method allows to capture the global structure of the data even with a small batch size, which is beneficial for large-scale real-world data. We show that our method captures the structure of the manifold more accurately than the recently introduced topological autoencoder (TopoAE).

Author Information

Anastasiia Varava (KTH Royal Institute of Technology)
Danica Kragic (KTH Royal Institute of Technology)
Simon Schönenberger (ETH Zurich)
Jen Jen Chung (KTH Royal Institute of Technology )
Roland Siegwart (ETH Zurich )
Vladislav Polianskii (KTH Royal Institute of Technology )

More from the Same Authors