Gloria

Results 28 issues of Gloria

I use Blenderproc to generate datasets, however, it version is blender 3.1 however, `bpy.context.scene.render.alpha_mode = 'TRANSPARENT' # in ['TRANSPARENT', 'SKY']` the code in your nocs_map_cube.py, this code is not support...

When I run the test.py ```python import tsnecuda testcuda.test() ``` its occour it's problem. Do you know how to solve it problem. ![image](https://github.com/CannyLab/tsne-cuda/assets/39433014/6fa212b3-238f-43c0-84dd-da7b12cebff8)

I the datasets download. But I find it's format dose not meet your code ```python IMAGE_DIRECTORY = "./dataset/potsdam/train_img_512_1_ori" LABEL_DIRECTORY = "./dataset/potsdam/train_lab_512_ori" IMAGE_DIRECTORY_TARGET = "./data/vaihingen/train_img_512" LABEL_DIRECTORY_TARGET = "./data/vaihingen/train_lab_512" VAL_IMAGE_DIRECTORY_TARGET = "./data/vaihingen/val_img_512"...

```txt Download the pre-trained [p2v_best_m.pth](https://github.com/RS-CSU/MemoryAdaptNet-master/blob/main) and put it in the snapshots folder ``` I click the link of the pre-trained model,But 404.I think something wrong with your link.

![image](https://github.com/lhoyer/DAFormer/assets/39433014/383bd7e9-ae95-4abb-a77d-0ed655f02126) I try use T-sne to visual the features. But I meet some problems. There are a total of 500 verified images on the citycapes dataset. I use the model...