Dynamic Scenes Data Set
Christoph Feichtenhofer1, Axel Pinz1 and Richard P. Wildes2
1 Institute of Electrical Measurement and Measurement Signal Processing
Graz Univesity of Technology
2 Department of Electrical Engineering and Computer Science
and Centre for Vision Research
York University, Toronto, ON, Canada
provides a dynamic scenes video data
set. This new data set samples from 20 scene classes, while encompassing a wide range of conditions, including those arising from natural within scene category differences, seasonal and diurnal variations as well as viewing parameters. Thumbnail examples of each class are shown below. For each scene class in the dataset, there are 60 colour videos, with no two samples for a given class taken from the same physical scene. Half of the videso within each class are acquired with a static camera and half are acquired with a moving camera, with camera motions encompassing pan, tilt, zoom and jitter. Having both static and moving camera instances for each class allows for systematic consideration of the role this variable plays in scene categoriation. All videos have been compressed with H.264 codec using the ffmpeg video ibrary. Duration for each video is 5 seconds, with orignal frame rates ranging between 24 and 30 frames per second. All have been resized to a maximum widhth of 480 pixels, while preserving their original aspect ratio.
frames of all scenes from the YUP++
Dynamic Scenes data set.
of our technical approach to
dynamic scene understanding, see our project
(YUP++) Dynamic Scenes Dataset
Old version: (YUPENN) Dynamic Scenes Dataset
C. Feichtenhofer, A. Pinz and R.P. Wildes, Temporal residual networks for dynamic scene recognition,
Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017.
C. Feichtenhofer, A. Pinz and R.P. Wildes, Dynamic scene recognition with complementary spatiotemporal features,
IEEE Transactions on Pattern Analysis and
Machine Intelligence (PAMI), 2016.
C. Feichtenhofer, A. Pinz and R.P. Wildes, Bags of spacetime energies for dynamic scene recognition,
Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2014.
C. Feichtenhofer, A. Pinz and R.P. Wildes, Spacetime forests with complementary features for dynamic scene recognition,
Proceedings of the British Machine Vision Conference (BMVC), 2013.
Derpanis, M. Lecce, K. Daniildis and R.P.
Dynamic Scene Understanding: The Role of
Orientation Features in Space and Time
in Scene Classification, IEEE Conference Computer Vision and
Pattern Recognition (CVPR), 2012.
updated: May 19, 2017