Inspiration

Decentralized AI (DeCoAI) can mitigate the risk of manipulation by centralized entities in the development and use of AI models. This is important because AI models can have a massive impact on people's lives, and centralized entities may have incentives to manipulate these models for their own benefit.

What it does

DeCoAI is a project that allows users to upload images and receive a classification result indicating whether the image is a cat, dog, or other. The project uses a pre-trained machine learning model from TensorFlow to generate the classification result.

How we built it

Using Machine learning model weights from tensorflow for the classification problem of cats and dogs generate the output and integrate with AWS instance with the frontend provided by hypercycle

Challenges we ran into

The team behind DeCoAI had difficulty finding a free and open-source text-to-image generation library. This is a relatively new and challenging area of research. However, there are a number of commercial text-to-image generation libraries available.

Accomplishments that we're proud of

The team behind DeCoAI is proud of having completed a working end-to-end AI computation on an AWS instance. This is a significant accomplishment, as it demonstrates the feasibility of using DeCoAI to provide AI services to users.

What we learned

The team behind DeCoAI learned how to run classification algorithms, integrate with AWS instances, and use remote SSH for real-time collaboration. These are all valuable skills for developing and deploying AI applications.

What's next for DeCoAI

The team behind DeCoAI plans to integrate a blockchain-based subscription model and wallet-based authentication into the platform. This would make it possible for users to pay for DeCoAI services using cryptocurrency. The team also plans to develop other AI services, such as text-to-image generation, custom photo editing, and video editing.

Share this project:

Updates