Zobrazeno 1 - 10
of 21
pro vyhledávání: '"Maynord, Michael"'
We introduce LEAP (illustrated in Figure 1), a novel method for generating video-grounded action programs through use of a Large Language Model (LLM). These action programs represent the motoric, perceptual, and structural aspects of action, and cons
Externí odkaz:
http://arxiv.org/abs/2312.00055
In this paper we introduce a rule-based, compositional, and hierarchical modeling of action using Therbligs as our atoms. Introducing these atoms provides us with a consistent, expressive, contact-centered representation of action. Over the atoms we
Externí odkaz:
http://arxiv.org/abs/2304.03631
Autor:
Dessalene, Eadom, Devaraj, Chinmaya, Maynord, Michael, Fermuller, Cornelia, Aloimonos, Yiannis
Human actions involving hand manipulations are structured according to the making and breaking of hand-object contact, and human visual understanding of action is reliant on anticipation of contact as is demonstrated by pioneering work in cognitive s
Externí odkaz:
http://arxiv.org/abs/2102.00649
Autor:
Dessalene, Eadom, Maynord, Michael, Devaraj, Chinmaya, Fermuller, Cornelia, Aloimonos, Yiannis
We introduce Egocentric Object Manipulation Graphs (Ego-OMG) - a novel representation for activity modeling and anticipation of near future actions integrating three components: 1) semantic temporal structure of activities, 2) short-term dynamics, an
Externí odkaz:
http://arxiv.org/abs/2006.03201
We introduce Evenly Cascaded convolutional Network (ECN), a neural network taking inspiration from the cascade algorithm of wavelet analysis. ECN employs two feature streams - a low-level and high-level steam. At each layer these streams interact, su
Externí odkaz:
http://arxiv.org/abs/1807.00456
Akademický článek
Tento výsledek nelze pro nepřihlášené uživatele zobrazit.
K zobrazení výsledku je třeba se přihlásit.
K zobrazení výsledku je třeba se přihlásit.
Autor:
Chen, Weijie, Astley, Susan M., Farhangi, Mehdi, Maynord, Michael, Sahiner, Berkman, Petrick, Nicholas
Publikováno v:
Proceedings of SPIE; April 2024, Vol. 12927 Issue: 1 p129272S-129272S-5, 1163454p
Akademický článek
Tento výsledek nelze pro nepřihlášené uživatele zobrazit.
K zobrazení výsledku je třeba se přihlásit.
K zobrazení výsledku je třeba se přihlásit.
Autor:
Fermüller, Cornelia, Maynord, Michael
Publikováno v:
In Advanced Methods and Deep Learning in Computer Vision 2022:373-403
Autor:
Farhangi, Mohammad Mehdi, Maynord, Michael, Fermüller, Cornelia, Aloimonos, Yiannis, Sahiner, Berkman, Petrick, Nicholas
Publikováno v:
Journal of Medical Imaging; Jul/Aug2024, Vol. 11 Issue 4, p44507-44507, 1p