Transforming Rooms with Generative Technology

Project Summary:

Built an automated system using generative AI that could generate the interior design for input image of the room based on a given prompt.


1. Add the required interior at a specific location without changing the room structure.

2. Process the prompt and create an object which can be suitable as per the room.


Technical Challenges:

1. To find the edges of the room that can be used to manage the room structure as it is.

2. Establishing ETL pipeline to take in real-time data for real-time prediction for interior design.

3. Minimize the inference time to process and generate suitable interior design.


The Solution That We Proposed:

An interactive & centralized web application enabled with generative AI models that can generate attractive & suitable interior design as per the user's demand. Prepare the model to generate designs for any type of room like a living room, bedroom, any type of office, etc. without knowledge of prompt engineering. The user can write only the category of the room and the AI system can generate the best suitable interior on its own.


AI models were to be made in a way that they keep self-learning even after deployment to make sure they adapt to the scenarios in real-time. Integrate all the above functionality and AI models with desktop applications, With high efficiency.


Measurable Benefits Post Implementation:

With the usage of AI models & user-friendly Web applications, clients benefited in the following ways:


1. Using this accurate end-to-end system generates a creative idea about the interior in a single click and with minimal time.

2. The client can use this automated AI system for freemium plans in their product.