A multi-task learning based hybrid prediction algorithm for privacy preserving human activity recognition framework
چکیده
There is ever increasing need to use computer vision devices capture videos as part of many real-world applications. However, invading privacy people the cause concern. for protecting while are used purposefully based on objective functions. One such case human activity recognition without disclosing identity. In this paper, we proposed a multi-task learning hybrid prediction algorithm (MTL-HPA) towards realising preserving framework (PPHARF). It serves purpose by recognizing activities from identity humans present in multimedia object. Face any person video anonymized preserve actions exposed get them extracted. Without losing utility recognition, anonymization achieved. Humans and face detection methods file reveal persons video. We experimentally confirm with joint-annotated motion data base (JHMDB) daily action localization YouTube (DALY) datasets that recognises ensures non-disclosure information. Our approach better than traditional techniques noise adding, blurring, masking.