Dynamic scene understanding: The role of orientation features in space and time in scene classification

By Konstantinos G. Derpanis, Matthieu Lecce, Kostas Daniilidis, Richard P. Wildes. 2012

129 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

C3D: Generic Features for Video Analysis (2014, methodology): “Our ﬁnding is, C3D consistently outperforms all other 2D image deep features on large-and medium-size datasets: Sport1M, UCF101, ASLAN, and YUPENN.”
Geometry Guided Convolutional Neural Networks for Self-Supervised Video Representation Learning (2018, methodology): “Particulary, our approach achieves 17.7% better on the Maryland dataset and 10.2% better on the YUPENN dataset than the second best selfsupervised representation learning approaches [24].”
Dynamic Scene Classification: Learning Motion Descriptors with Slow Features Analysis (2013, methodology): “This motion feature uses Histograms of Optic Flow (HOF) in a similar spirit to the static images features SIFT [22] or HOG [7].”
Spacetime Forests with Complementary Features for Dynamic Scene Recognition (2013, methodology): “During training, the temporal alignment of the video slices is treated as latent; correspondingly, each temporal slice of each training video generates its own feature vector according to procedures of Sec.”
Learning Spatiotemporal Features with 3D Convolutional Networks (2014, methodology): “YUPENN consists of 420 videos of 14 scene categories and Maryland has 130 videos of 13 scene categories.”
Seeing the Arrow of Time (2014, methodology): “As in the prior work, we split each video into 2 × 2 spatial sub-regions, and concatenated the SOE responses for each sub-region for form a final feature.”
Application of Machine Learning in Wireless Networks: Key Techniques and Open Issues (2018, methodology): “For cache management, authors in [81] utilize a mixture of data set YUPENN [137] and UFC101 [137] as their”
Bags of Spacetime Energies for Dynamic Scene Recognition (2014, background): “The dominant approach to pooling throughout the dynamic scene recognition literature is that of spatial pyramids [6, 9, 22, 24], which employ predefined hierarchically gridded aggregation regions to capture coarse geometric layout in a multiresolution fashion.”
Self-Supervised Spatio-Temporal Representation Learning for Videos by Predicting Motion and Appearance Statistics (2019, methodology): “We follow the same leave-one-out evaluation protocol as described in [8].”