Zobrazeno 1 - 10
of 19
pro vyhledávání: '"Saleh, Fatemeh Sadat"'
Autor:
Ben-Shabat, Yizhak, Yu, Xin, Saleh, Fatemeh Sadat, Campbell, Dylan, Rodriguez-Opazo, Cristian, Li, Hongdong, Gould, Stephen
The availability of a large labeled dataset is a key requirement for applying deep learning methods to solve various computer vision tasks. In the context of understanding human activities, existing public datasets, while large in size, are often lim
Externí odkaz:
http://arxiv.org/abs/2007.00394
Autor:
Zhang, Jing, Fan, Deng-Ping, Dai, Yuchao, Anwar, Saeed, Saleh, Fatemeh Sadat, Zhang, Tong, Barnes, Nick
In this paper, we propose the first framework (UCNet) to employ uncertainty for RGB-D saliency detection by learning from the data labeling process. Existing RGB-D saliency detection methods treat the saliency detection task as a point estimation pro
Externí odkaz:
http://arxiv.org/abs/2004.05763
Autor:
Aliakbarian, Sadegh, Saleh, Fatemeh Sadat, Petersson, Lars, Gould, Stephen, Salzmann, Mathieu
We tackle the task of diverse 3D human motion prediction, that is, forecasting multiple plausible future 3D poses given a sequence of observed 3D poses. In this context, a popular approach consists of using a Conditional Variational Autoencoder (CVAE
Externí odkaz:
http://arxiv.org/abs/1912.08521
Autor:
Rodriguez-Opazo, Cristian, Marrese-Taylor, Edison, Saleh, Fatemeh Sadat, Li, Hongdong, Gould, Stephen
This paper studies the problem of temporal moment localization in a long untrimmed video using natural language as the query. Given an untrimmed video and a sentence as the query, the goal is to determine the starting, and the ending, of the relevant
Externí odkaz:
http://arxiv.org/abs/1908.07236
Autor:
Aliakbarian, Mohammad Sadegh, Saleh, Fatemeh Sadat, Salzmann, Mathieu, Petersson, Lars, Gould, Stephen, Habibian, Amirhossein
Human motion prediction is a stochastic process: Given an observed sequence of poses, multiple future motions are plausible. Existing approaches to modeling this stochasticity typically combine a random noise vector with information about the previou
Externí odkaz:
http://arxiv.org/abs/1908.00733
Autor:
Aliakbarian, Mohammad Sadegh, Saleh, Fatemeh Sadat, Salzmann, Mathieu, Fernando, Basura, Petersson, Lars, Andersson, Lars
Action anticipation is critical in scenarios where one needs to react before the action is finalized. This is, for instance, the case in automated driving, where a car needs to, e.g., avoid hitting pedestrians and respect traffic lights. While soluti
Externí odkaz:
http://arxiv.org/abs/1810.09044
Autor:
Saleh, Fatemeh Sadat, Aliakbarian, Mohammad Sadegh, Salzmann, Mathieu, Petersson, Lars, Alvarez, Jose M.
Training a deep network to perform semantic segmentation requires large amounts of labeled data. To alleviate the manual effort of annotating real images, researchers have investigated the use of synthetic data, which can be labeled automatically. Un
Externí odkaz:
http://arxiv.org/abs/1807.06132
Autor:
Saleh, Fatemeh Sadat, Aliakbarian, Mohammad Sadegh, Salzmann, Mathieu, Petersson, Lars, Alvarez, Jose M.
Pixel-level annotations are expensive and time-consuming to obtain. Hence, weak supervision using only image tags could have a significant impact in semantic segmentation. Recent years have seen great progress in weakly-supervised semantic segmentati
Externí odkaz:
http://arxiv.org/abs/1708.04400
Autor:
Saleh, Fatemeh Sadat, Aliakbarian, Mohammad Sadegh, Salzmann, Mathieu, Petersson, Lars, Alvarez, Jose M., Gould, Stephen
Pixel-level annotations are expensive and time consuming to obtain. Hence, weak supervision using only image tags could have a significant impact in semantic segmentation. Recently, CNN-based methods have proposed to fine-tune pre-trained networks us
Externí odkaz:
http://arxiv.org/abs/1706.02189
Autor:
Aliakbarian, Mohammad Sadegh, Saleh, Fatemeh Sadat, Salzmann, Mathieu, Fernando, Basura, Petersson, Lars, Andersson, Lars
In contrast to the widely studied problem of recognizing an action given a complete sequence, action anticipation aims to identify the action from only partially available videos. As such, it is therefore key to the success of computer vision applica
Externí odkaz:
http://arxiv.org/abs/1703.07023