Developed by OARC at UCLA in collaboration with SDSC and TACC
Object Detection
We'll be building an app to find everyday objects in a photograph. You give it an image; it gives you back the same image with boxes drawn around what it found, plus a machine-readable list of those objects, how confident it was about each one, and where in the picture they are.
It uses a general-purpose model trained to recognize eighty common categories — people, vehicles, animals, furniture, food, everyday household items. It cannot find things outside that list, so a picture with none of them in it correctly comes back empty.
This is the batch version: it runs once, does its work, and exits. There is nothing to click and nothing to watch.
Creation steps (streamlined for tutorial walk-through)
1. Navigate to the left-hand menu -> Create -> App
You will be taken to the "Create App" form, where you can fill in the following values.
| Form name | Value |
|---|---|
| App Name | <username> YOLO11 |
| App ID | <username>.apps.tutorial.object_detection.batch.expanse |
| App version | 0.1.0 |
| Summary | Object Detection |
| App type | Batch |
| Container runtime | Singularity |
| Singularity runtime options | Singularity run |
| Scheduler profile | Add existing -> start typing “expanse” -> select and add "Expanse scheduler profile for ospdev" |
| Container image name | /home/bwinjum/singularity/object-detection/app-object-detection-sif_latest.sif |
| Available system | Expanse ospdev |
| Default system | Expanse ospdev |
| Input type | Form |
| Parameters -> Container launch arguments | --bind ${PWD}:/data |
| Runtime resources -> Memory | 6000 MBs |
When all the above look ok, click "Save" at the bottom -> you should get a banner notification that it was created and you can now configure the form.
2. Configure the app form
We will be adding just one new element to the runtime resources you should already see listed.
a) At the top, move the selected tab name from View -> Build
b) select Add Element
c) select File (OSP) -> Add element and fill in the following
| Form name | Value |
|---|---|
| Title | Image |
| Type | Command line argument |
| Command prefix | --image |
| Add trailing space to command prefix | Checked |
| Add trailing space to command suffix | Checked |
| Form validation | Select Required Required message: Image file is required |
| File Settings -> Allowed file extensions | jpg png |
d) when finished, select "Save"
3. Test run
At the top, move the selection from Build back to View
Upload your image in the form and use Allocation: sds277
Then click on Submit!
4. Monitor job and View results
After submitting, the status page shows the job's progress.
- Possible statuses include Staging, Processing inputs, Running, Finished, Failed.
- You can terminate a running or queued job from this page by clicking Terminate Job
- Click Output to review results after completion.