Noise Removal is a system that deals with the obstacles appears in the video. While capturing a video there may come unknown obstacles that may degrade the video. Repeating of the video time and again is time consuming and is not possible all the time. People may not record the same clip twice, nor can get back the video without that obstacle. So, the noise content video can create big problem by damaging its quality. SiamMask data is used in this project so that it directly predicts the mask of an object. Deep CNN algorithm has been used in this project which takes the video as an input, then the user selects the portion which they want to remove from the video the video is reconstructed and the output is made noise free. It removes the obstacles and replace it with the similar background. Hence, this relies on the clip of before and after of the selected obstacle. This application can be implicated for all the people to get the noise free video.