ActionProjects
ActionProjects
Activity specific motion models for activity recognition:c
Activity Specific Motion Models:
Activity Repetition Recogniton:
Temporal Feature Weighting for Prototype-Based Action Recognition
Hierarchical NMF for Detection
Temporal Feature Weighting for Prototype-Based Action Recognition
Efficient Human Action Recognition and Detection
We present a human action recognition system suitable
for very short sequences. In particular, we estimate Histograms
of Oriented Gradients (HOGs) for the current
frame as well as the corresponding dense flow field estimated
from two frames. The thus obtained descriptors are
then efficiently represented by the coefficients of a Nonnegative
Matrix Factorization (NMF). To further speed up
the overall process, we apply an efficient cascaded Linear
Discriminant Analysis (CLDA) classifier. In the experimental
results we show the benefits of the proposed approach
on standard benchmark datasets as well as on more challenging
and realistic videos. In addition, since other stateof-
the-art methods apply weighting between different cues,
we provide a detailed analysis of the importance of weighting
for action recognition and show that weighting is not
necessarily required for the given task.
Publications:
Videos:
Activity Specific Motion Models:
Activity Repetition Recogniton:
Temporal Feature Weighting for Prototype-Based Action Recognition
Hierarchical NMF for Detection
Temporal Feature Weighting for Prototype-Based Action Recognition
Efficient Human Action Recognition and Detection
We present a human action recognition system suitable
for very short sequences. In particular, we estimate Histograms
of Oriented Gradients (HOGs) for the current
frame as well as the corresponding dense flow field estimated
from two frames. The thus obtained descriptors are
then efficiently represented by the coefficients of a Nonnegative
Matrix Factorization (NMF). To further speed up
the overall process, we apply an efficient cascaded Linear
Discriminant Analysis (CLDA) classifier. In the experimental
results we show the benefits of the proposed approach
on standard benchmark datasets as well as on more challenging
and realistic videos. In addition, since other stateof-
the-art methods apply weighting between different cues,
we provide a detailed analysis of the importance of weighting
for action recognition and show that weighting is not
necessarily required for the given task.
Publications:
- Mauthner T., Roth P. M. and Bischof, H. (2009): Action Recognition from a Small Number of Frames, Proceedings of Computer Vision Winter Workshop, 2009
- Mauthner T., Roth P. M. and Bischof, H. (2009): Instant Action Recognition , Proceedings of Scandinavian Conference on Image Analysis, 2009-June
- Roth P. M., Mauthner T. and Bischof, H. (2009): Efficient Human Action Recognition by Cascaded Linear Classification , Proceedings of 1st IEEE Workshop on Video-Oriented Object and Event Classification during ICCV09, 2009-October
Videos:
| Download | Detection results for human actions in realistic hand-held videos trained from Weizmann database. In addition sports videos, using controlled indoor environments for training, and detecting during competition. Both applications show how robust and generalizing the proposed methods are. |
Document Actions
