Ignacio Martinez-Alpiste, Gelayol Golcarenarenji, Qi Wang and Jose M. Alcaraz Calero
"DynPedestrian" deployes an adaptative solution to YOLO-based models for autonomous driving. It is able to detect pedestrians at different distances and it is prepared to work on low-consumption NVIDIA gpu. This software can achieve a speed increment of 48.7% without dropping the accuracy.
International Journal publication.