We got our inspiration from personal experience. Resul Ucar has been editing videos for over 5 years and the task he hates the most is the simple cut. To edit videos you have to spend time watching the whole video, cut the parts you want in and delete the parts you don't. This may sound simple to do but when you have to do this for 5 hours worth of footage it gets a bit repetitive. We used the power of AI and automation to make a project that can save you hours on your video editing.
What it does
Our program lets the user set keywords as cut points in the video which the program then listens for and uses to edit out the unwanted parts of the video and then create a completed video by combining the good takes which will be represented by speaking a different keyword at the end of the take.
How we built it
Our program takes in recorded videos and uses IBM Watson’s speech to text functionality to locate keywords that will mark out key points in the videos. After locating the points of the keywords, it then cuts the video down and concatenates all the good clips into a completed video. We used Python, IBM Watson, FFMPEG and Linux to develop the program.
Challenges I ran into
We wanted to use Google Cloud’s Speech to text API but we encountered bugs and issues which made it unsuitable for our purposes, therefore we switched to IBM Watson.
Accomplishments that I'm proud of
We are proud of how compact the logic of the speech processing portion of our program was.
What I learned
We learned a lot about using Google Cloud Services even though in the end we decided not to use it. We also learned a lot about Python as we have little to no experience in this programming language.
What's next for Snip's and Clip's
We hope to expand this software out to other platforms like Windows and macOS because it currently only works on Linux. Also, we would like to implement a GUI to make it easier to use.