Deep Learning Algorithms for Human Activity Recognition in Manual Material Handling Tasks.

In the 70-30 split training, the BiLSTM and RCNN performed similarly, but the RCNN showed an increasing trend of the F1-score as the number of subjects in the dataset increased, suggesting that it could result in higher classification performances with more subjects.

also in 63,971 other papers

In this case, the average F1-score shows a slightly increasing trend with increasing subject numbers. 4.3.

also in 801 other papers

RCNN performed similarly (95.9%; 89.2%) with a positive trend with subject numerosity.

also in 11,005 other papers