DeepForest
DeepForest copied to clipboard
weecology
Python Package for Airborne RGB machine learning
**Describe the bug** Hello, I am trying to process a GeoTIFF using DeepForest and the release model. when I run the package it returns with "no predictions made, returning none"...
**Please describe A feature you would like to be added.** I'm building examples of existing models for tree crown mapping and publishing them in the EnvDS book. **Describe the solution...
**Describe the bug** shapefile_to_annotations shown on bird_detector example does not have the shown arguments: 'box_points' and 'buffer_size' Does the method just need to updated on the Git repo? **To Reproduce**...
Hi everyone, first of all, I'd like to thank the authors for all of their work on _deepforest_ which is what got me into _deep learning_. As a student, I've...
**Describe the bug** I'm not sure if this is a bug or my ignorance, but would really appreciate the help. Whenever I attempt to train the release model regardless of...
This issue is meant for cataloging data that should eventually go into a deepforest baseline. Happy to have contributions from the community on this issue. https://arxiv.org/pdf/2208.10607.pdf https://github.com/jonathanventura/urban-tree-detection-data A portion of...
1. Profile existing dataloader. 2. Test whether preload_images=True should happen before create a dataloader 3. Have a performance page, timings with tests across multiple gpu If you preload in the...
``` def create_model(num_classes, nms_thresh, score_thresh, backbone = None): """Create a retinanet model Args: num_classes (int): number of classes in the model nms_thresh (float): non-max suppression threshold for intersection-over-union [0,1] score_thresh...
you can pass kwargs to main.create_trainer(strategy==...), this replaces the distributed_backend argument, #342 following: https://pytorch-lightning.readthedocs.io/en/latest/extensions/strategy.html