Presentation
The MUMTDB DataBase can be used for evaluating recognition systems under the multi-user situation.
The handwriting data is collected based on two predefined mind map scenarios which contains 9 different classes of graphical symbols. Each scenario is completed by involving two users at the same time.
Since the users are given freedom to draw the symbols as they want, the dataset contains a diversity of multi-stroke and even multi-touch symbols. It allows addressing new challenging problems regarding the recognition of simultaneous composition of structured documents. A total of 20 people participated in the data acquisition and were partitioned half to half into two groups. Users of group 1 were asked to draw the symbols on a 27” touch screen by fingers. They were given all the freedom to draw the symbol in any way or order they prefer in order to record the way people naturally draw. Since users in this group draw symbols by fingers, some users tend to draw the symbols like ’bracket’ and ’double line’ and others with two fingers in the same time. In other words, this dataset contains not only multi-stroke but also multi-touch symbols. Users from group 2 made the acquisition on a 80” touch screen by stylus. In this case they have more free space to draw the symbol and may simultaneously draw different symbols very close to each other. Since all the symbols are drawn by stylus, the multi-touch case does not exist in this group.
Note that each pair of users were asked to exchange their positions to do the acquisition twice. Finally, a total of 40 diagrams are obtained in our dataset.
The dataset is freely available.We would like to thank the Excence company from their contribution in the data collection and their financial support of the project.
The diagram data acquisition procedure on a 80” touch screen. Two users are drawing the diagram together using styluses.
Ground truth of a mind map diagram:
Download link
Contact
For any question please contact Eric Anquetil.
