AI deep face changing video evidence obtaining method based on pulse signals

The invention discloses an AI deep face-changing video evidence obtaining method based on pulse signals, which is characterized in that cardiovascular pulse waves propagated in a human body periodically cause vascular wall extension, so that the light absorption capability of tissues containing a large number of blood vessels is also synchronously fluctuated, and regular pulse signals are reflected. In the face video shooting process, the micro changes which cannot be seen by naked eyes can be recorded by a common camera, and the regular micro changes can be destroyed by the forged face generated by the AI method. According to the characteristic, a classifier obtained through training by combining a machine learning algorithm SVM is combined, and the pulse signals due to abnormal human bodies in a vivid face forged video represented by deep learning are effectively recognized, and therefore the purpose of video evidence obtaining is achieved. According to the method, the video to be detected does not need to