What is an AI growing up video, and how do you turn old photos into a video of someone growing up?

An AI growing up video is a video built from still photos of one person across their life that shows them growing up in continuous motion. At one they are crawling, at four they are running, at thirteen they are on a stage, at forty they are hugging their own kids. Every one of those transitions is generated by AI, but the start and end point of each transition is one of your real photos.

It is not a slideshow. In a slideshow the photos stay frozen and swap with a fade. In a growing video there is no transition between photos, there is motion: the photo itself becomes the next one. People searching for an AI photo to video growing up tool, or a way to animate old photos of a grandparent, usually mean this. The full comparison is in slideshow vs AI growing video.

The technique is called image to video generation with a start frame and an end frame. In plain words: you give the model two photos and tell it "begin here and finish there". The model generates everything in between.

1. Put the photos in age order. Photo 1 is the earliest, the last photo is from today.
2. Each consecutive pair becomes one job. Photo 1 and photo 2, photo 2 and photo 3, and so on.
3. The model gets photo N as the first frame and photo N+1 as the last frame, along with an instruction about the kind of motion, and generates about five seconds of video in which the person moves, turns or walks, changing age along the way.
4. The clips join with no seam. Each clip starts on exactly the photo where the previous one ended, so no cut or fade is needed.
5. An opening, music and titles are added and everything is compiled into one video file.

That is also where the math comes from: N photos give N minus 1 clips. On Memvivo, 13 photos make one minute of video, and each extra minute needs 12 more photos. Details in how many photos you need for a video.

Three kinds of tools sound alike but give very different results:

A classic photo morph video of someone aging moves points on the face and blends pixels, so it looks like one face melting into another. Single photo animation brings one moment to life but does not move through time. On Memvivo that is essentially Classic style, which suits sets with different people in each photo. The growing video, Growing style, is the default: it generates full body motion within a scene and carries it from one age to the next.

• Clear faces looking at the camera. The model needs to see the face to keep the identity.
• The same person in every photo, or the same couple. Growing style works with two people in frame and keeps each identity.
• Chronological order. A photo in the wrong place looks like a jump backwards in age.
• Reasonable age gaps between neighbouring photos. Age 3 to age 6 works well. Age 3 straight to age 30 is much harder.
• Good light and sharpness. Old or scanned photos get automatic face restoration before animation, but a good source always wins. See how to scan old photos with your phone.
• Similar angles and distances. Two close-up portraits join more smoothly than a close-up and a distant shot against scenery.

Go in with the right expectations. AI generates motion that looks believable, not a record of what really happened. And there are situations it finds hard:

• Identity drift. When two photos were taken from very different angles, say a profile and a straight-on portrait, the face in the middle of the clip can briefly look like someone else.
• Extreme age gaps. Going straight from a baby to an adult asks the model to invent too much. In-between photos solve most of it.
• Low resolution. A tiny or blurry photo gives the model little to work with. Face restoration helps, but it does not perform miracles.
• Many people in frame. A couple works well. In a group photo it is hard for the model to know who turns into whom.
• Hands, text and small objects sometimes come out distorted, as with any video model today.

These are family photos, often of children, so it is worth asking a few questions before uploading to any service: where the photos are stored, who they are sent to, whether they are used to train models, and how to delete them.

On Memvivo the answers are simple. Photos are stored in your account only, sent to our video providers solely to generate the clips, and deleted there within 14 days. They are never used to train models. You can delete the whole project from the dashboard at any time, and it disappears from our servers too. One more thing: if other people appear in the photos, like partners or friends, it is fair to ask them before sharing the video publicly.

• Bar and bat mitzvahs. Birth to thirteen, played at the venue. See bar and bat mitzvah videos.
• Milestone birthdays. 60 for Mom, 70 for Dad, 80 for Grandpa.
• Anniversaries. Two people growing side by side, from the first date to the grandchildren.
• Retirements, tributes and family gatherings. The story of someone the whole family loves. See tribute videos.
• A baby's first year. One photo a month, and a short video where you watch them grow.

1. Try it free. The first animation is free, no credit card. Upload two photos of the same person at different ages and watch the transition.
2. Upload all your photos in order. From your phone or computer, up to 80 per project. Photos of album pages work too.
3. Wait a few minutes. Clips start generating while you upload, and each one shows a live estimate.
4. Review the transitions. Regenerate any you do not like with a different motion.
5. Choose an opening, music and titles. A cinematic opening of about five seconds with their name, your own song, one from the library or an original song, a closing card and a caption per clip.
6. Download a 1080p MP4 in 16:9 for screens and TVs, 1:1 for Facebook, or 9:16 for Stories and Reels.

Payment is one time by video length, no subscription, with a 30 day full refund from your account settings. Current prices are on the pricing page.