Startup ideas💡: Replacing faces service

 Startup ideas💡, [03.08.21 10:47]

Restaurant feedback

Some restaurants, cafes and so switched(or switching) to an online menu, online ordering while sitting in a place. It simplifies a whole workflow for customers and restaurant's workers.

Customers don't need to wait to order. If a place is busy, it may be difficult to "catch" a waiter and make an order. Waiters don't need to check all the customers: who needs to make an order, who needs to pay, etc.

This idea is about adding feedback functionality that asks customers what they think about a place after they paid the check. Or, what they think no matter of the payment.

A bigger idea of it is creating software to support all the workflow: an app to make an order, summon a waiter/administrator, leave feedback, pay, etc.

Startup ideas💡, [08.08.21 19:20]

Cloud video transformation

Many companies build their own solutuions to work with videos. For instance, encoding, decoding, cutting, making previews and so on. For a company which goal isn't providing such functionality, it's resource consuming to make its own solutions to work with video.

The idea is to provide video transformations functionality as API in a cloud. Thus, I can easily, for example, cut a video file, then reduce its size in one or two API calls instead of spending a huge amount of time implementing this myself.

Monetization: users pay for an API call. Or tiers for a specific amount of calls per day, week, or month.

Startup ideas💡 

Replacing faces service

My friend makes videos where he substitutes people's faces. He tried a few paid services, but there are problems: the video quality is low, the faces aren't replaced right.

So, the main issues are the quality of the video and the substitution. We might build a service that provides a better video quality which means higher processing time on a server. To leverage this issue, we could provide custom plans where users would choose video quality. For example, 720p is priced more than 480p. Notice that it's hard to render such videos and probably you'll need performant GPUs for cloud hosting.

The second issue is the quality of the face substitution. To handle this we could use custom parameters provided to end-users but simplified to their understanding. So the service can be used by not only programmers, video professionals, or deep learning engineers. Our target is regular people that aren't familiar with underhood face replacement magic.

There are some open-source projects for that, so probably you don't need to re-create a wheel. I guess it needs a custom configuration to handle various kinds of videos and faces to be handled in the service and to provide a friendly API for users.