Timezone: »
Zero-Shot Learning (ZSL) aims at classifying unlabeled objects by leveraging auxiliary knowledge, such as semantic representations. A limitation of previous approaches is that only intrinsic properties of objects, e.g. their visual appearance, are taken into account while their context, e.g. the surrounding objects in the image, is ignored. Following the intuitive principle that objects tend to be found in certain contexts but not others, we propose a new and challenging approach, context-aware ZSL, that leverages semantic representations in a new way to model the conditional likelihood of an object to appear in a given context. Finally, through extensive experiments conducted on Visual Genome, we show that contextual information can substantially improve the standard ZSL approach and is robust to unbalanced classes.
Author Information
Eloi Zablocki (Sorbonne Université, LIP6)
Patrick Bordes (Laboratoire d'Informatique de PARIS VI)
Laure Soulier (Sorbonne Université)
Benjamin Piwowarski (Sorbonne Université)
Patrick Gallinari (LIP6, Sorbonne Universite)
Related Events (a corresponding poster, oral, or spotlight)
-
2019 Poster: Context-Aware Zero-Shot Learning for Object Recognition »
Thu. Jun 13th 01:30 -- 04:00 AM Room Pacific Ballroom #131
More from the Same Authors
-
2022 : Building a Subspace of Policies for Scalable Continual Learning »
Jean-Baptiste Gaya · Thang Doan · Lucas Caccia · Laure Soulier · Ludovic Denoyer · Roberta Raileanu -
2020 Poster: Stochastic Latent Residual Video Prediction »
Jean-Yves Franceschi · Edouard Delasalles · Mickael Chen · Sylvain Lamprier · Patrick Gallinari -
2020 Poster: Discriminative Adversarial Search for Abstractive Summarization »
Thomas Scialom · Paul-Alexis Dray · Sylvain Lamprier · Benjamin Piwowarski · Jacopo Staiano