Perspective of Homogenous Motion Reincarnation Based on Artificial Intelligence

Provided by: International Journal for Development of Computer Science & Technology (IJDCST)
Topic: Software
Format: PDF
Learning-based approaches for homogenous motion reincarnation often rely on large training sets. Most of these approaches do not perform well when only a few training samples are available. In this paper, the authors consider the problem of homogenous motion reincarnation from a multi clip per action. Each clip contains at most 25 frames. Here in this paper, they are using a novel technique of image reincarnation based on artificial intelligence. Algorithm for extracting and classifying multi clip motion in a sequence based on motion trajectories.

Find By Topic