Rémi Cresson
Rémi Cresson
Hi @akanimax First of all many thanks for this amazing contribution. In the paper, it is stated `we modified the gradient penalty to be the average of the penalties over...
More documentation related to the use of docker should ease greatly user experience! Here is a list of stuff that could be very useful to put in the documentation: -...
OTBTF 3.0 will make use of TensorFlow v2 APIs (python and C++) - [x] use TF 2.0 libraries - [x] use v1 models with ```tf.disable_v2_behavior()``` - [ ] port models...
Currently, model outputs of same size are handled in inference. # Outputs parameters A nice enhancement could be to add a **parameter group for models outputs**, and use the same...
The beginners-friendly `TensorflowModelTrain`application does not implement the RMSE metric yet
Provide `TileHint` at patches extraction filters output.
The use of `outlabels` is confusing when working on something different than single-valued patch-based classification/regression problems. It could be better to have a more generic framework, that relies solely on...
Develop an application to apply object detection models on rasters, and generate output vector data with bounding boxes.
It could be great to have this somewhere. Basically it could consist in the following steps: - a combination of OTB `orthorectification` (or `gdalwarp`), to prepare: - one LR image...