TU Darmstadt / ULB / TUbiblio

Diverse Image Captioning with Context-Object Split Latent Spaces

Mahajan, Shweta ; Roth, Stefan (2020)
Diverse Image Captioning with Context-Object Split Latent Spaces.
34th Conference on Neural Information Processing Systems (NeurIPS 2020). virtual Conference (06.-12.12.2020)
Konferenzveröffentlichung, Bibliographie

Kurzbeschreibung (Abstract)

Diverse image captioning models aim to learn one-to-many mappings that are innate to cross-domain datasets, such as of images and texts. Current methods for this task are based on generative latent variable models, eg. VAEs with structured latent spaces. Yet, the amount of multimodality captured by prior work is limited to that of the paired training data -- the true diversity of the underlying generative process is not fully captured. To address this limitation, we leverage the contextual descriptions in the dataset that explain similar contexts in different visual scenes. To this end, we introduce a novel factorization of the latent space, termed context-object split, to model diversity in contextual descriptions across images and texts within the dataset. Our framework not only enables diverse captioning through context-based pseudo supervision, but extends this to images with novel objects and without paired captions in the training data. We evaluate our COS-CVAE approach on the standard COCO dataset and on the held-out COCO dataset consisting of images with novel objects, showing significant gains in accuracy and diversity.

Typ des Eintrags: Konferenzveröffentlichung
Erschienen: 2020
Autor(en): Mahajan, Shweta ; Roth, Stefan
Art des Eintrags: Bibliographie
Titel: Diverse Image Captioning with Context-Object Split Latent Spaces
Sprache: Englisch
Publikationsjahr: 6 Dezember 2020
Verlag: Curran Associates
Reihe: Advances in Neural Information Processing Systems
Band einer Reihe: 33
Veranstaltungstitel: 34th Conference on Neural Information Processing Systems (NeurIPS 2020)
Veranstaltungsort: virtual Conference
Veranstaltungsdatum: 06.-12.12.2020
URL / URN: https://proceedings.neurips.cc/paper/2020
Zugehörige Links:
Kurzbeschreibung (Abstract):

Diverse image captioning models aim to learn one-to-many mappings that are innate to cross-domain datasets, such as of images and texts. Current methods for this task are based on generative latent variable models, eg. VAEs with structured latent spaces. Yet, the amount of multimodality captured by prior work is limited to that of the paired training data -- the true diversity of the underlying generative process is not fully captured. To address this limitation, we leverage the contextual descriptions in the dataset that explain similar contexts in different visual scenes. To this end, we introduce a novel factorization of the latent space, termed context-object split, to model diversity in contextual descriptions across images and texts within the dataset. Our framework not only enables diverse captioning through context-based pseudo supervision, but extends this to images with novel objects and without paired captions in the training data. We evaluate our COS-CVAE approach on the standard COCO dataset and on the held-out COCO dataset consisting of images with novel objects, showing significant gains in accuracy and diversity.

Fachbereich(e)/-gebiet(e): 20 Fachbereich Informatik
20 Fachbereich Informatik > Visuelle Inferenz
DFG-Graduiertenkollegs
DFG-Graduiertenkollegs > Graduiertenkolleg 1994 Adaptive Informationsaufbereitung aus heterogenen Quellen
Hinterlegungsdatum: 07 Mär 2022 09:52
Letzte Änderung: 07 Mär 2022 09:52
PPN:
Zugehörige Links:
Export:
Suche nach Titel in: TUfind oder in Google
Frage zum Eintrag Frage zum Eintrag

Optionen (nur für Redakteure)
Redaktionelle Details anzeigen Redaktionelle Details anzeigen