Hybrid Deep Artificial Humming Bird Algorithm For Improved Real Time Blind Assistance with Advanced Jetson Nano GPU

Blindness often known as the inability for vision, Blindness continues to be a serious problem despite several innovations. Existing Blind assistance systems are bulky, more Costly and not effective in real time environment for priority based object detection. So proposed Hybrid deep Artificial Humming Bird Algorithm for improved real time Blind Assistance. This model implemented using SUN- RGB -D dataset in Spider environment using Python and real time input processed using High speed light weight Jetson Nano GPU. Priority of objects determined using rank base criteria. Proposed model shows best results and obtained accuracy 92.07 %, F1 Score 90.08 % and frame processing time 0.13 s at batch size sixty.