A scalable solution for colorizing black and white videos using DeOldify and AWS is introduced in the following text. The solution utilizes AWS services to create an efficient workflow capable of handling multiple videos at once and producing impressive colorization results.

A Scalable Video Colorization Solution

TrackIt has developed a workflow centered around DeOldify, a powerful Open Source deep learning model for colorizing black and white images and videos. The solution also leverages AWS services such as Amazon S3, AWS Lambda, Amazon Elastic Transcoder, Amazon SQS, and Amazon EC2 Spot Instances. The workflow provides the following benefits:

  1. Process multiple videos simultaneously: Video segmentation is used to handle multiple video clips at one time. The parallel processing significantly reduces the time required to colorize large batches of videos.
  1. Optimize costs and performance: Amazon EC2 Spot Instances and Auto Scaling Groups are utilized to cut costs while ensuring optimal performance for video colorization.
  1. Benefit from a fully managed and integrated solution: AWS services are fully integrated and a seamless and efficient colorization process is enabled without manually managing resources.

A Scalable Video Colorization Solution with DeOldify and AWS

Workflow Overview

  1. A black and white video is uploaded to an Amazon S3 input bucket
  2. An AWS Lambda function is triggered which segments the video using Amazon Elastic Transcoder
  3. The video segments are stored in another S3 bucket and a message is sent to an Amazon SQS queue
  4. Amazon EC2 Spot Instances managed by an Auto Scaling Group are launched to process the video segments
  5. The colorized video segments are stored in an output S3 bucket
  6. Another Lambda function is triggered to merge the colorized segments back into a single video file
image
Colorization Workflow AWS Architecture Diagram

Cost Breakdown

Below are the estimated workflow costs for a 1-hour video with both On-Demand and Spot instances using g4dn.xlarge instances in the Oregon region:

Total cost of the workflow for a 1-hour video:

  • Amazon S3 storage:
    • Input bucket: $0.023 (assuming 10GB storage)
    • Output bucket: $0.023 (assuming 10GB storage)
    • Total: $0.046
  • Amazon S3 data transfer: $0.18 (assuming 40GB outbound data transfer)
  • Amazon Elastic Transcoder: $0.90 (assuming 60 minutes of video transcoding)
  • AWS Lambda:
    • Lambda for segmenting: $0.001
    • Lambda for merging: $0.001
    • Total: $0.002
  • Amazon SQS: $0.004 (assuming 4,000 requests)
  • Amazon EC2 On-Demand (g4dn.xlarge):
    • 3 hours * $0.526/hour = $1.578
  • Amazon EC2 Spot (g4dn.xlarge):
    • The cost will vary but assuming a 70% discount on the on-demand rate:
    • 3 hours * $0.1578/hour = $0.4734

Total On-Demand cost: $2.71

Total Spot cost: $1.61

Note: These costs are estimates based on the mentioned assumptions and the AWS rates at the time of writing. Actual costs may vary depending on usage, processing duration, and updated AWS rates.

The Stunning Results of Colorization

High-quality colorization results breathe new life into older black and white content. Below is an example of a video that has been colorized using the aforementioned process:

Before / After Colorization –  Excerpt from The Pawnshop, Charlie Chaplin

Conclusion

Creating a video colorization farm with DeOldify and AWS with integrated video segmentation can be an efficient and cost effective solution. It enables the processing of multiple videos simultaneously and in an optimized manner using AWS services to deliver impressive colorization results. This video colorization processing solution showcases TrackIt’s deep technical expertise in AWS and understanding of media workflows. Companies interested in leveraging this scalable video colorization solution can contact TrackIt for more information.

About TrackIt

TrackIt is an international AWS cloud consulting, systems integration, and software development firm headquartered in Marina del Rey, CA.

We have built our reputation on helping media companies architect and implement cost-effective, reliable, and scalable Media & Entertainment workflows in the cloud. These include streaming and on-demand video solutions, media asset management, and archiving, incorporating the latest AI technology to build bespoke media solutions tailored to customer requirements.

Cloud-native software development is at the foundation of what we do. We specialize in Application Modernization, Containerization, Infrastructure as Code and event-driven serverless architectures by leveraging the latest AWS services. Along with our Managed Services offerings which provide 24/7 cloud infrastructure maintenance and support, we are able to provide complete solutions for the media industry.