Deep learning from temporal coherence in video
Proceedings of the 26th annual international conference on machine learning, 2009•dl.acm.org
This work proposes a learning method for deep architectures that takes advantage of
sequential data, in particular from the temporal coherence that naturally exists in unlabeled
video recordings. That is, two successive frames are likely to contain the same object or
objects. This coherence is used as a supervisory signal over the unlabeled data, and is used
to improve the performance on a supervised task of interest. We demonstrate the
effectiveness of this method on some pose invariant object and face recognition tasks.
sequential data, in particular from the temporal coherence that naturally exists in unlabeled
video recordings. That is, two successive frames are likely to contain the same object or
objects. This coherence is used as a supervisory signal over the unlabeled data, and is used
to improve the performance on a supervised task of interest. We demonstrate the
effectiveness of this method on some pose invariant object and face recognition tasks.
This work proposes a learning method for deep architectures that takes advantage of sequential data, in particular from the temporal coherence that naturally exists in unlabeled video recordings. That is, two successive frames are likely to contain the same object or objects. This coherence is used as a supervisory signal over the unlabeled data, and is used to improve the performance on a supervised task of interest. We demonstrate the effectiveness of this method on some pose invariant object and face recognition tasks.
ACM Digital Library以上显示的是最相近的搜索结果。 查看全部搜索结果