One face in a whole graduation recording.
This 55-second demonstration follows the class speaker in Belmont High School’s 2002 graduation, from a long ceremony recording to an excerpt of his personal reel.
Start with a videoWhat the demonstration shows
The source recording lasts about 2 hours and 37 minutes, and the scan returned 159 face groups. The video above is an edited explanation, not the entire source, the complete output reel or a real-time recording of a scan. The interface animations illustrate the steps.
- The ceremony recording runs quickly while the scan and face grouping are illustrated.
- The class speaker is selected from the faces found in the recording.
- Five chosen moments lead into a vertical reel, with a blurred backdrop and music underneath the original voices.
The five moments were checked
The chosen appearances show the speaker during his speech and at the podium later in the ceremony. They come from two camera views.
The initial scan also grouped a choir singer with the speaker. For this public demonstration, the five chosen moments were checked by hand and those mismatches were left out. This is why reviewing the preview matters: face matching can miss or mix people, especially in distant stage shots.
Try the same workflow with your recording
Make graduation video highlights Find a specific person in a long video
- Upload an MP4, MOV or WebM file up to 15 GB, or paste a YouTube link to a video you own or have permission to use.
- Confirm your email with a one-time code — there’s no password. Your upload keeps going while you do.
- KinCut scans the video on its own GPU and groups every face it finds. We email you when it’s done.
- Name the people you recognise and pick who belongs in the reel. KinCut gathers every moment they appear.
- Watch a free preview of the highlight reel and short clips. Unlock full quality when you’re happy.
Source footage and privacy
The ceremony was recorded by Belmont Community Television on 9 June 2002 and is available on Internet Archive under the Public Domain Mark. The demonstration uses Sunflower Valley, a CC0 music track, alongside original recorded speech.
KinCut searches for the faces you select within your own recording. It does not identify an unknown graduate against an external database. This is public demonstration material, not footage taken from a customer’s account.
Last updated: