Alan R Lowe

Seems like: + Cellpose https://github.com/MouseLand/cellpose or + Usiigaci (MaskRCNN) https://github.com/oist/Usiigaci might be a good places to start. @KristinaUlicna - shall we add a notebook showing how to go from the...

> This is awesome. We are having a student to work on porting Usiigaci's result to work with bayesian tracker. super impressive. Exciting! Please let us know if we can...

Hi Mark, You should be able to use the `assign_class_ID` flag in the [segmentation_to_objects](https://btrack.readthedocs.io/en/latest/api/btrack.utils.segmentation_to_objects.html#btrack.utils.segmentation_to_objects) function to do this. From the tracks you can then get the original mask labels using:...

Hi Peter, It looks like things are working. For example, if I look at cell 45 - it divides into 175 and 176, and correctly builds the tree (I can...

Ah - I see, OK. Well this can happen during the optimisation step for a couple of reasons: * if there are competing hypotheses that account for the segmentation *...

Hi @Surya059 - we hadn't really envisioned `btrack` being used on resource constrained devices. Do you have a use case for this?

Indeed, this happens when you're generating too many hypotheses per track. It becomes difficult to opimize. Usually, this is because the configuration is not set up properly for your dataset....

Also, you can set a timeout for the optimizer by passing in some options, e.g.: ```python tracker.optimize(options={'tm_lim': int(6e4)}) ``` This will stop the python script from freezing for a long...

Hi @samueljrosen, Yes, it can be a bit tricky on M1. I've found the following to work: - Start a clean conda environment - Install napari first, using their [guidelines](https://github.com/napari/napari#installation)...

Hi Ciro, Haven't had much time to work on this unfortunately. Is there a reason why you need the GPU accelerated version? Related #39