Experimental data --- Robotic Untangling of Herbs and Salads with Parallel Gripper

This data supports the work done in the paper titled "Robotic Untangling of Herbs and Salads with Parallel Gripper" accepted for publication in the Proceedings of the 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2020). Readme file is provided.<br><br>The data folder consists of 5 csv files. Files follow the below naming convention.<br> "purpose"_"material".csv <br> A). train_plastc.csv -- training data as collected picking plastic herbs. B). test_plastic.csv -- testing data as collected picking plastic herbs using predicted gripper width. C). train_wild_rocket.csv -- training data as collected picking wild rocket. D). test_wild_rocket.csv -- testing data as collected picking wild rocket using predicted gripper width. E). test_parsley.csv -- testing data as collected picking flat leaf parsley using predicted gripper width. <br> <br>Column details -----------------<br>1. train data<br> columns: gripper_width(mm) -- represents the gripper aperture width in mm. weight_picked(g) -- represents the weight picked in grams. type -- represents the method used. "Normal" and "Extended" represents GI and spread-and-pick method. 2. test data<br> columns: weight_picked_GI(g) -- represents the picked weight following GI method. weight_picked_Spread_Pick(g) -- represents the picked weight following spread and pick method. target_weight(g) -- represents the target weight. predicted_griper_width(mm) -- represents the predicted gripper width as used for picking. <br>