Throwing destination generation algorithm training Dataset
This two files are dataset of throwing handover between two person. It contains 420 samples of 7 people. One of them is the gesture data of catcher before throwing and the other is for after throwing. They have the body and hands position and the score for throwing quality. This files are used to train the destination generation algorithm.
As a data collecting and capturing tool, the same system of experiment is employed except that the thrower and catcher both are in very random conditions by their own decision and the sensor’s spatial position is fixed to a unique reference on thrower’s body. This system is used for developing dataset for training section of system and also for process of collecting body gesture data of catcher for using the destination position generation system
This two files are dataset of throwing handover between two person. One of them is the gesture data of catcher before throwing and the other is for after throwing. 420 samples of 7 people with different height and ages were collected. The dataset of every sample contains 27 data which are listed in below. For optimal system training, it is necessary to preprocess and deliver them properly to the network training system.
Before Throwing data: 1- Catcher's ID , 2- Catching try , 3- Geature of body , 4- Catcher's Body (X), 5- Catcher's Body (Y), 6- Catcher's Body (Z), 7- Catcher's Left Hand (X), 8- Catcher's Left Hand (Y), 9- Catcher's Left Hand (Z), 10- Catcher's Right Hand (X), 11- Catcher's Right Hand (Y), 12- Catcher's Right Hand (Z), 13- Direction of the body, 14- Height
After Throwing data: 15- Catcher's ID , 16- Catching try , 17- Catching Position (X) , 18- Catching Position (Y), 19- Catching Position (Z), 20- Catcher's Body (X), 21- Catcher's Body (Y), 22- Catcher's Body (Z), 23- Success of Catching, 24- Ease of catching Evaluated by Catcher, 25- Calculated Hardness, 26- Direction of the body in Degree after Catching, 27- Catching Hand