Major Updates to our Future Work Research Items

[nleadholm]

We’re excited to announce a series of significant updates to our research-focused “Future Work” items. If you would like to tackle one of these yourself, you can find out more by exploring them in the tool available on our website.

To highlight a few examples:

• Implement alternative methods for Monty to extract depth information
• Explore performance on a larger dataset of objects with 2D + 3D compositional arrangements
• Design a replicable real-world test bed, for example by leveraging 3D printed and open-source components
• Implement a 1D sensor module

If you were already familiar with the older list of Future Work items, you can get an overview of the changes by looking at the PRs listed below:

#1050

#1076

#1084

[nleadholm]

@jmwright your project here could be a great starting point for the third item I highlighted!

Results With a Low-Cost Stereo Camera Rig

I did an experiment to see if I could train Monty on CAD models, then 3D print those models, and then get Monty to recognize the prints on a low-cost stereo rig that I’ve built. It was a simple experiment with only three objects and a static rig (no turntable): A mug A glass that was just the mug with the handle removed A block that is essentially a large calibration cube [stereo_rig_with_test_prints] The glass and the mug were speckled using stick-on labels. Bare glossy filament gives a…

[jmwright]

Thanks. Will that task require an RFC if the stereo rig will work with the unmodified Monty main branch?

I’ve got several questions about what TBP would want to see in that hardware demo, but I’m not sure where to post them. This thread doesn’t seem like the right place to be asking lower-level questions.

[nleadholm]

Great question, no, and in general that item doesn’t require an RFC because nothing would be merged into tbp.monty. As you suggest, it would still make sense to discuss any questions you have early on about the implementation approach. A new Topic along the lines of “Replicable real-world test bed: 3D printing + stereo rig” on here would be a perfect place to talk about this more.