Deep Learning for Activity Recognition Using Audio and Video

Neural networks have established themselves as powerhouses in what concerns several types of detection, ranging from human activities to their emotions. Several analysis exist, and the most popular successful is video. However, there are other kinds analysis, which, despite not being used often, still promising. In this article, a comparison between audio video drawn an attempt classify violence detection real-time streams. This study, which followed CRISP-DM methodology, made use models available through PyTorch order test diverse set achieve robust results. The results obtained proved why has such prevalence, with classification handily outperforming its counterpart. Whilst attained on average 76% accuracy, secured scores 89%, showing significant difference performance. study concluded that applied methods quite promising detecting violence, using both