NEAXT - New Generation Enabling Art Technology

NEAXT app provides opportunities to Artists and Galleries to scout and promote artistry talents all across the globe powered by Artificial Intelligence & Blockchain.

Comment

Inspiration

  1. Covid has changed the world including Art industry. It was reported that marketing loss - income of 75-80% in museums, global decrease of tourism - impacted heavily on ticket sells.
  2. 70% of the museums increased their online presence measures. There's an increase in online visits to museum in this new normal.
  3. Demand of art and artists is still there and statistically even more than before but it's the connection gap that is causing loss in revenue.

There's a strong need of a platform that helps art galleries to get more audience and scout artists at global level to generate new trends and interests. Artists have been equally impacted in these covid times and hence there is a need for a solution that empowers them digitally so that they generate better revenue and traction for their art work.

What it does

NEAXT - New Generation Enabling Art Technology

An Artificial Intelligence powered Personalised Art RecommendationaI system helps Art Galleries and Auction Houses to go global and online by generating curiosity and demand among art seekers and admirers to visit their showrooms to view and buy art work of various Artists using various collaborative filtering tecniques. NEAXT also helps Artists to publish their art work online for users to buy, sell, rent, own fractional ownership of the artwork using Blockchain technology. NEAXT also helps the new and younger generation to gain interest about Art by learning through various Podcasts & Quick shorts uploaded by artists sharing tutorials and various interesting facts about region & art culture.

How we built it

We have created a responsive web application that works on any device be it Mobile, Tablet, Laptop, etc. by using emerging technologies - AI and Blockchain, frontend technologies - React, JavaScript and backend technologies - Python, Flask

We have build an Artificial Intelligence model that helps in personalised art recommendation as per user preference. Artworks, for example, paintings, are not the same as ordinary items for sale on the Internet: they have their own style, content, and emotion: simply using tags for recommendation is a compromise to pale reality. We have built a recommendation system based on visual recognition. Art delivers a feeling, and we want to use deep neural networks to explain the impressions delivered by the object's visual appearances. We use two neural networks to extract the features such as color, shape, edge, etc., from the paintings and obtain feature vectors. After dimension reduction, we calculate the distance between these feature vectors and find the closest ones for the recommendation. This system works amazingly well; it can show users the painting with similar content, style, and emotion. When users browse paintings on our platform, we can always know what they want to see next.

Tags are vital for human to distinguish the art. To extract the information from artwork images, a mobile-drivable neural network is deployed, this customized CNN model extracts the physical content from the artwork images such as information related to art styles, object content, color, etc.

Challenges we ran into

At first, we were getting difficulties in integration Artificial Intelligence module with our Mobile / Web Application but post building multiple modules, we managed to build a flask (python based backend) and were able to host the model and connect it via frontend application.

Due to timing constraints, we haven't been able to connect our Artificial Intelligence Model with Blockchain. Currently, they are separate module but we are working on integrating it so that every single actions from all modules will be publicly available on blockchain.

Accomplishments that we're proud of

We successfully developed working individual components such as frontend, backend, recommendation model, auto-tagging model and blockchain module.

What we learned

To have clear understanding among team members how individual components will integrate with each other (inputs and outputs) for smooth integration and less/no re-work

What's next for NEAXT

Currently, we are using local blockchain but will deploy it on a public blockchain whenever we get a chance for developing it as POC for few Artists, Art galleries and Auction Houses.

Built With

Share this project:

Updates