Posted on

Scott Status Report #1 (2/16)

  • What did you personally accomplish this week on the project?
    • Research into relevant AWS Services and similar projects online for reference.
      • Found research online about using ec2 instance for running openpose
      • https://michaelsobrepera.com/guides/openposeaws.html
      • Came up with cloud architecture that should allow for our capabilities on AWS services
        • Camera video -> AWS Kinesis -> Openpose on ec2 -> AWS Sagemaker -> Lambda (serverless) -> API Gateway
    • Talked with professor Nace about getting credits for AWS, also in progress of discussing how we can use our $600 budget for AWS services.
    • Initial research into output data of openpose to use for our classifier
      • https://github.com/CMU-Perceptual-Computing-Lab/openpose/blob/maste r/doc/output.md
    • Local installation of OpenPose
  •  Is your progress on schedule or behind?
    • We are on schedule. At this point we have ordered the hardware we will need and done enough preliminary research that we feel our general architecture will be successful. This is up to our expectations for the week.
  • What deliverables do you hope to complete in the next week?
    • In the next week we will hopefully receive our hardware and be able to begin the implementation. I want to first set up AWS Kinesis on the pi so that we can get the video stream in the cloud. The next goal after this would be to install openpose on AWS which could be tricky, but after this is done we should be able to see live openpose data in the cloud from our camera.