SERVICE DETAILS
PoseNet TF Body Pose Service (by AI4EOSC)
Overview
OSCAR exposed service that runs the AI4EOSC PoseNet TensorFlow container and publishes its DEEPaaS API through OSCAR. Users can access the proxied OpenAPI UI and invoke body-pose inference methods directly from the browser or any HTTP client.
Deployment requirements
- Memory
- 4 GiB
- Processors
- 2 vCPU0 GPU
Software
Files & artifacts
- fdl.yml (opens in a new tab)text/yaml
Service Definition (FDL)
Show content of fdl.yml
- script.sh (opens in a new tab)text/x-shellscript
Service Execution Script
Show content of script.sh
- README.md (opens in a new tab)text/markdown
Service Notes
Show content of README.md
- icon.png (opens in a new tab)image/png
Service Icon
Show content of icon.png
- 001.jpg (opens in a new tab)image/jpeg
Sample body pose image — JPEG image used to validate the exposed PoseNet prediction endpoint.
Show content of 001.jpg
- Expected prediction ZIPExpected artifactapplication/zip
Expected prediction ZIP — ZIP archive returned by the PoseNet predict endpoint.
Usage & acceptance tests
Exposed PoseNet ZIP acceptance test
Upload the sample image to the exposed DEEPaaS predict endpoint using service Basic Auth and verify that the response is a valid ZIP archive.
Invoke the exposed PoseNet predict endpoint with the sample image and verify that the response is a ZIP archive.
RO-Crate metadata
- application/json
Show content of ro-crate-metadata.json