What is the A.I. Apps Ecosystem?
Unleash live’s unique A.I. Apps ecosystem brings Industry leaders and AI developers together to solve real world problems. The A.I. App ecosystem is designed to give developers and industry leaders the opportunity to search for relevant AI solutions for implementation and development.
Find an A.I. App that suits you
- Knowing exactly what you want to solve is the key to getting the most from A.I. Apps. The first stage is understanding what the critical things you want to identify in video or images to improve your decision making process.
- Search the A.I. App store to find an applicable app that will provide you with the insights you need to make faster decisions.
- Once you have found the A.I. App of choice select the tile it expand the description, select the Try or Buy button and follow the prompts to activate it. Repeat this process if you decide to deactivate an A.I. App on your account.
Selecting TRY will allow a short trial period for the A.I. App, test it out to see if it suits your needs. Selecting BUY will activate your access to the A.I. App allowing you to apply the algorithm to library stored media and live streams.
If you are a developer how do you use the A.I. App Ecosystem?
It's a simple framework to create, upload and monetise your industry relevant artificial intelligence apps. All you need is an idea and a good feel for your potential customers and demand.
1. To begin you will need to sign up to a Unleash live developer account, simply enable the developer account from within the Unleash live profile page.
2. Now the coding begins. Using tf object_detection framework you code the model for training. Training your models will require sufficient video footage and/or images for accuracy. There are 2 ways of gaining this footage/images: Either through your own resources or we can connect you with our customer base.
Our friends at Google provide a fantastic resource through Colab to train your models. Check it out, it's free for 12 hours of GPU training.
3. Once you are happy with the model accuracy it is time to test. Your model graphs are frozen for deployment. As developer you then upload the num classes / pb / pbtxt into the Unleash live developer sandbox.
The code is tested against videos and images stored in our Unleash live developer Sandbox or other footage you may obtain.
4. When the required accuracy has been achieved during the testing phase you submit the model for approval. The submission will require an estimated price, description of the app, the problem it solves and up to 6 images of the app UI (for display). See see our submission form framework for all the details.
5. The model is then tested with our Unleash live testing team, amendments are requested and when it is cleared by our team your A.I. model is published and available for purchase by Unleash live customers.
Finally, improvements and updates can be regularly committed to your developer Sandbox.