Výsledky vyhledávání - "DOĞAN, Pelin"

Report

Enriching Video Captions With Contextual Text

Autor: Rimle, Philipp, Dogan, Pelin, Gross, Markus

Understanding video content and generating caption with context is an important and challenging task. Unlike prior methods that typically attempt to generate generic video captions without context, our architecture contextualizes captioning by infusi

Externí odkaz: http://arxiv.org/abs/2007.14682

Zobrazit plný text záznamu

Akademický článek

Tento výsledek nelze pro nepřihlášené uživatele zobrazit.
K zobrazení výsledku je třeba se přihlásit.

Report

Neural Sequential Phrase Grounding (SeqGROUND)

Autor: Dogan, Pelin, Sigal, Leonid, Gross, Markus

We propose an end-to-end approach for phrase grounding in images. Unlike prior methods that typically attempt to ground each phrase independently by building an image-text embedding, our architecture formulates grounding of multiple phrases as a sequ

Externí odkaz: http://arxiv.org/abs/1903.07669

Zobrazit plný text záznamu

Report

A Neural Multi-sequence Alignment TeCHnique (NeuMATCH)

Autor: Dogan, Pelin, Li, Boyang, Sigal, Leonid, Gross, Markus

The alignment of heterogeneous sequential data (video to text) is an important and challenging problem. Standard techniques for this task, including Dynamic Time Warping (DTW) and Conditional Random Fields (CRFs), suffer from inherent drawbacks. Main

Externí odkaz: http://arxiv.org/abs/1803.00057

Zobrazit plný text záznamu