EchoLense

Inspiration

As students new to AWS, we wanted to create a meaningful project that took advantages of the AWS services while having a simple architecture.

What it does

EchoLense lets users upload an image and receive a detailed, spoken description of the scene

How we built it

We used AWS S3 for storage, Lambda for processing, and Amazon Bedrock to generate image captions. A simple frontend connects everything.

Challenges we ran into

Creating a lambda function We were able to solve this issue by watching youtube demonstrations and using large language models to help us. Connecting the S3 bucket with our lambda function For a long time, we were able to hardcode the image into the lambda function, but couldn’t figure out how to connect the s3 bucket with the function to produce an output. We were able to solve this by making small changes to the working version and adapting it to the s3 configuration until it worked. Connecting the lambda output API back with our output A large issue we had with connecting the API with our application was the CORS configurations. We were able to figure out the configurations with help of AI.

Accomplishments that we're proud of

Being new to AWS, we were overwhelmed with the knowledge it took to use all of these tools. We're proud that we were able to execute a working version of our idea that can be used to help visually impaired individuals.