Face Animation in a Photo: From Still Portrait to Emotion
Face animation adds movement to a portrait: the person smiles, blinks, looks aside or nods slightly. The key is to preserve recognisability and avoid turning the face into a mask.
What the Neural Network Animates
- Expression. Smile, raised brows, a hint of surprise.
- Gaze. Eye movement and looking direction.
- Blinking. Natural eyelid movement.
- Head movement. A slight tilt or nod.
- Lips. Articulation for a greeting video.
The model does not replace the person — it adds motion on top of the existing portrait. That is why the source photo should be sharp and large enough.
Motion Scenarios
For a gift: a soft smile, blinking and a slight nod. For social media: more active expression and a head turn. For a historical portrait: minimal motion to preserve the formal mood.
If you are making a video with text, record the voice or choose the music first, then match the animation length.
Mistakes That Ruin the Result
- Too wide a smile on a photo where the person looks serious.
- Blurry eyes and small details — the model cannot animate them.
- Sharp head turns when the source is a profile.
- Shadows and highlights that stick to the face as it moves.
Compare the result with the original: if the person no longer looks like themselves, reduce the motion and return to a more neutral portrait.
Try Face Animation
Upload a portrait and choose motion. The neural network will create a short video you can download and use in personal projects.