Skip to content

How to implement other models? #2

Description

@Proeliorr

The article mentions:

(...)
We trained four models: one pan-European model trained on all training blocks (six in total, two from each country), and three biome-specific models for boreal, temperate, and Mediterranean forests separately (trained on two blocks of data from the relevant biome)
(...)

which gives the assumption that both the pan-European model and the three biome-specific variants would be included to allow users to choose the most appropriate model for their application.

The readme.md has in its content:

Model Selection
Within the model folder, we have biome-specific as well as more general ecosystem-agnostic models.

Although after pulling the repo (I have the git-lfs preinstalled), only the eu.pth archive is downloaded. Should there be any other files available? - or is there any other parameter which I did not find?

The example running line from the manual has a parameter --model, and uses eu.pth.
python3 predict.py --point-cloud /x/x/cloud.ply --model eu.pth --batch_size 16 --is-wood 0.55 --grid_size 2.0 --min_pts 512;

The manual does not provide any information about extracting the eu.pth file, which suggests an inconsistency in the data availability.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions