DEV Community

Cover image for We created a new Image Segmentation model, seeking your feedback
Fannie Lin
Fannie Lin

Posted on

We created a new Image Segmentation model, seeking your feedback

We are building an API solution for computer vision to make model development easier ~

Even though we have been aware that a lot of repetitive work can be replaced by computer vision much more efficiently, ❌not all companies have started their transformation journey.

👉This could be due to the following reasons:

  1. Model development is expensive in general.
  2. Talent crunch.
  3. Lack of training data, etc.

💪We want to change this scenario, to make it more accessible:

  1. with no need for professional AI knowledge,
  2. with less training data
  3. and less training time, than ever.

Thus, we built this API product, which enables users to build their computer vision models in Minutes. For example, to segment ships with only 4 data points and 52 seconds of training like this: https://www.youtube.com/watch?v=2SFMvUk9WYk

And another example is to build a 🧢PPE model (people protection equipment detection), here's how:
① Annotate 5 data points;
② 🚀Wait for 46 seconds for training.
③💣 Your model is ready to go, integrable, and scalable.

The inference result is shown below:
Image description

🚀We are about to launch our beta product and thus are seeking early users to help us iterate and improve! The whole building process is usually less than 10 Minutes per model including data annotation.

😊If you are willing to help us, please leave your info here (https://www.ailiverse.com/waiting-list) and we will provide you with access to our documentation and GitHub repo soon! Thank you in advance! ♥

Top comments (2)

Collapse
 
newbieai profile image
NewbieAI

I am in, happy to try!

Collapse
 
fannieailiverse profile image
Fannie Lin

Thanks friend! Will get back to you soon!