Tempo Temporal Forgery Video Detection Using Machine Learning Approach

This research paper explore a new way of detecting video forging between frames and intraframes by referring to the correlation coefficient using a frame continuity relationship. For any given video set, a groundbreaking technique called the "Spatio Temporal copy" produces video forgery detection using a machine learning method based on the continuous correlation between the conjugative sequences and the group of frames from the forgery video. The proposed forgery detection algorithm aims to identify the sequence groups forged intermediately by referring to the SVM classifier. Changing in the video sequence can result in a different fingerprint than collected initially, either at the spatial or at the temporal levels. Awareness of the statistical features that add frame continuity is the foundation for developing our algorithm to identify the video forgery detection that creates the duplicate. In the sequential continuity of the forged structures, we successfully identified the copy-move and copy delete frames, combining spatial and temporal fingerprints in an orderly and systematic approach. By referring the forensic standard data sets such as SULFA, VTD, and REWIND, we have tested and obtained high accuracy results with prominent researchers in the forensic video area