NIPS
1998
15 years 8 days ago
1998
The W-S (Wake-Sleep) algorithm is a simple learning rule for the models with hidden variables. It is shown that this algorithm can be applied to a factor analysis model which is a...
108
Voted
BMVC
1997
15 years 8 days ago
1997
In this paper we present an algorithm for structure and motion (SM) recovery under affine projection from video sequences. The algorithm tracks the motion of a single structure, b...
IRMA
2000
15 years 8 days ago
2000
The on-line algorithms in machine learning are intended to discover unknown function of the domain based on incremental observing of it instance by instance. These algorithms have...
IWMMDBMS
1998
15 years 8 days ago
1998
An algorithm for detecting scene changes in compressed video streams is proposed. The proposed algorithm directly exploits the motion compensation information and the prediction e...
ACL
1997
15 years 8 days ago
1997
The first step in most empirical work in multilingual NLP is to construct maps of the correspondence between texts and their translations (bitext maps). The Smooth Injective Map R...
75
Voted
IPCO
1998
15 years 8 days ago
1998
We introduce a gain-scaling technique for the generalized maximum ow problem. Using this technique, we present three simple and intuitive polynomial-time combinatorialalgorithms fo...
ICMAS
2000
15 years 9 days ago
2000
This paper addresses the issues involved in software agents making trade-offs during automated negotiations in which they have information uncertainty and resource limitations. In...
AAAI
1997
15 years 9 days ago
1997
The absolute loss is the absolute difference between the desired and predicted outcome. I demonstrateworst-case upper bounds on the absolute loss for the perceptron algorithm and ...
88
Voted
WSCG
2003
15 years 9 days ago
2003
We present a real-time rendering algorithm that generates soft shadows of dynamic scenes using a single light sample. As a depth-map algorithm it can handle arbitrary shadowed sur...
69
Voted
ESANN
2000
15 years 9 days ago
2000
: Recurrent neural networks possess interesting universal approximation capabilities, making them good candidates for time series modeling. Unfortunately, long term dependencies ar...