Metric learning aims to learn a highly discriminative model encouraging the embeddings of similar classes to be close in the chosen metrics and pushed apart for dissimilar ones. The common recipe is to use an encoder to extract embeddings and a distance-based loss function to match the representations - usually, the Euclidean distance is utilized. An emerging interest in learning hyperbolic data embeddings suggests that hyperbolic geometry can be beneficial for natural data. Following this line of work, we propose a new hyperbolic-based model for metric learning. At the core of our method is a vision transformer with output embeddings mapped to hyperbolic space. These embeddings are directly optimized using modified pairwise cross-entropy loss. We evaluate the proposed model with six different formulations on four datasets achieving the new state-of-the-art performance. The source code is available at https://github.com/htdt/hyp_metric.
Hyperbolic Vision Transformers: Combining Improvements in Metric Learning / Ermolov, Aleksandr; Mirvakhabova, Leyla; Khrulkov, Valentin; Sebe, Nicu; Oseledets, Ivan. - 2022-:(2022), pp. 7399-7409. (Intervento presentato al convegno 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2022 tenutosi a New Orleans nel 2022) [10.1109/CVPR52688.2022.00726].
Hyperbolic Vision Transformers: Combining Improvements in Metric Learning
Ermolov, Aleksandr;Sebe, Nicu;
2022-01-01
Abstract
Metric learning aims to learn a highly discriminative model encouraging the embeddings of similar classes to be close in the chosen metrics and pushed apart for dissimilar ones. The common recipe is to use an encoder to extract embeddings and a distance-based loss function to match the representations - usually, the Euclidean distance is utilized. An emerging interest in learning hyperbolic data embeddings suggests that hyperbolic geometry can be beneficial for natural data. Following this line of work, we propose a new hyperbolic-based model for metric learning. At the core of our method is a vision transformer with output embeddings mapped to hyperbolic space. These embeddings are directly optimized using modified pairwise cross-entropy loss. We evaluate the proposed model with six different formulations on four datasets achieving the new state-of-the-art performance. The source code is available at https://github.com/htdt/hyp_metric.File | Dimensione | Formato | |
---|---|---|---|
Ermolov_Hyperbolic_Vision_Transformers_Combining_Improvements_in_Metric_Learning_CVPR_2022_paper.pdf
accesso aperto
Tipologia:
Versione editoriale (Publisher’s layout)
Licenza:
Tutti i diritti riservati (All rights reserved)
Dimensione
2.77 MB
Formato
Adobe PDF
|
2.77 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione