Human Activity Recognition Using Deep Learning : A Survey

With the use of deep learning algorithms from artificial intelligence (AI), several types research have been conducted on video data. Object localization, behaviour analysis, scene understanding, labelling, human activity recognition (HAR), and event make up majority them. Among all them, HAR is one most difficult jobs key areas in data processing. can be used a variety fields, including robotics, human-computer interaction, surveillance, categorization. This seeks to compare approaches benchmark datasets for vision-based detection. We suggest brand-new taxonomy dividing literature into CNN- RNN-based methods. further categorise these four subgroups show methodologies, their effectiveness, experimental datasets. To illustrate development techniques, brief comparison also provided with handcrafted feature-based approach its merger learning. Finally, we go over potential future some unresolved issues recognising activities. survey's goal present recent developments techniques using analysis.