Policy to Recognize an Object Before Moving onto a New Object
Use the SM and/or LM goal generator to remain on an object until it is recognized (e.g. using basic image segmentation)
When there are multiple objects in the world (including different parts of a compositional object), it is beneficial to recognize the object currently observed (i.e. converge to high confidence) before moving onto a new object.
Such a policy could have different approaches, such as moving back if an LM believes it has moved onto a new object (reactive), or using the model of the most-likely-hypothesis to try to stay on the object (pro-active), or a mixture of these. In either case, these would be instances of model-based policies.
Initial prototyping has tested using a sudden decay of evidence for the most likely hypothesis as an indicator to move back into the object. However, this work was put on hold a while ago and needs to be picked up again.
Recently we have also added the ability for sensor modules to output goals. We could therefor also use model-free signals, such as goals based on segmentation of the input image, as a way to stay in a region with similar depth or features. This would help especially early on when the LMs don't have any strong hypotheses yet.
Help Us Make This Page Better
All our docs are open-source. If something is wrong or unclear, submit a PR to fix it!
Updated about 12 hours ago