The automatic generation of content continues to do its thing. The startup ZMO.ai which just announced an $8 million Series A fundraising round promises to automatically generate photo templates for businesses operating in the fashion industry.
AI at the service of fashion
This startup responds to a mission well known to fashion professionals. To promote its clothes, this industry needs models. These models will have to be made up, styled and then move on to dressing. Then we have the photo session which can take several hours. The startup's proposal is "simply" to generate virtual models.
In a world where fashion changes all the time (which isn't necessarily a good thing ecologically!), this part of the process is expensive. ZMO.ai therefore promises valuable time savings for companies. His proposal also lies in the fact that by reducing these costs, SMEs will be able to resist better against large companies.
100% bespoke mannequins
The use of GAN (generative-adversarial-Networks) makes it possible to create a custom-made entire human body with various parameters:
- Skin color
- The pose
- Various measurements
The creation of the model is done in four steps:
- Product photo upload
- Generate images 3D transparent mannequin
- Choose a model from a multitude of automatically generated models
- Download the obtained images
An important question remains: will this hyper-personalization lead to more realistic mannequins? Or will it lead to even more clichéd models coming ever closer to an unattainable ideal of beauty? The bets are open !
As for ZMO.ai, this company has a lot of ideas. Its fundraising will make it possible to finance other projects related to it. In particular, it plans to use GPT-3 to allow models to speak. We would then go from photo to video with promotional videos of models totally generated by artificial intelligence.