Data
A list of publicly available datasets for benchmarking of pattern recognition algorithms
Note that publicly available data is published in connection with the contests and competitions.
- Youtube 22 Concepts
- UCI Machine Learning Repository
- A list of datasets on the web by datawrangling.com
UCI standard database in a unified format
Some of the most popular UCI and Statlog datasets can be found in this directory, in a standard format, and split into standard partitions to make results more comparable. You will also find the code and script to make your own partitions - see the README file. (Thanks go to Roberto Paredes of the Universitat Politècnica de Valencia for this.)-
Hand-Written Symbol Recognition
Thanks go to Heloise Hse and A. Richard Newton of University California Berkely for this hand-written symbol database -
CAVIAR video sequences
The EC funded CAVIAR project (Context Aware Vision using Image-based Active Recognition) has collected and hand-labelled ground truth for 81 video sequences comprising about 90K frames. - Sequence Recognition Dataset
- MNist data from Yann LeCun
- USPS data from Max Planck Institute for Biological Cybernetics, Tübingen, Germany
- Dataset generator