Implement serverless semantic search of image and live video with Amazon Titan Multimodal Embeddings
Machine Learning Blog
This article introduces a serverless solution for performing semantic search on video and image data using Amazon Kinesis Video Streams, Amazon Bedrock's Titan Multimodal Embeddings model, and Amazon OpenSearch Service. The key points are:
Specifically, the article covers:
- Overview of the solution architecture combining Kinesis Video Streams, Titan Multimodal Embeddings, and OpenSearch Service
- How to optimize for functionality, accuracy, and cost by configuring frame extraction rates, embedding lengths, k-NN parameter, and leveraging serverless services like AWS Lambda
- Prerequisites and step-by-step instructions to deploy the solution using AWS Amplify
- How to use the deployed application for uploading files, capturing from webcam, and searching videos using text prompts
- Clean up steps to delete the deployed resources
- Conclusion highlighting the benefits and potential future applications
The AWS News Feed is currently looking for gold sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.
Related articles
Jun 6
2025
2025
Implement semantic video search using open source large vision models on Amazon SageMaker and Amazon OpenSearch Serverless
Apr 17
2026
2026
Power video semantic search with Amazon Nova Multimodal Embeddings
Feb 5
2025
2025
Video semantic search with AI on AWS
Nov 13
2024
2024
Build a reverse image search engine with Amazon Titan Multimodal Embeddings in Amazon Bedrock and AWS managed services
The AWS News Feed is currently looking for silver sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.