RozDavid / LanguageGroundedSemseg

Implementation for ECCV 2022 paper Language-Grounded Indoor 3D Semantic Segmentation in the Wild

RozDavid/LanguageGroundedSemseg Issues

Mapping CLASS_IDS to CLASS_NAME
Closed a year ago2
How to understand alpha in eqn.5?
Updated a year ago
Can I fetch individual objects from segmented point cloud?
Closed a year ago2
Is your code missing a processing step？
Updated 2 years ago7
Feeding segmentation labels to CLIP
Closed 2 years ago4
Logs of Language Grounded Pretraining
Closed 2 years ago1
RuntimeError: CUDA out of memory after many epochs
Closed 2 years ago1
Can't Test the Provided Pretrained Model Checkpoints
Closed 2 years ago3
Why this codebase require a lot of GPU Memory?
Closed 2 years ago1
Question about the environment
Closed 2 years ago
How to do the instance augmentation?
Closed 2 years ago1
Training time for pre-training and fine-tuning
Closed 2 years ago4
What the True format of test_output of instance segmentation batchmark?
Closed 2 years ago2
How can I get the predicted result?
Closed 2 years ago2
How to process the scannet200 data?
Closed 2 years ago3
Do you have estimated time for training on Scannet V2 from scratch?
Closed 2 years ago4
2D Visualizations
Closed 2 years ago6
Text-supervised Contrastive optimization.
Closed 2 years ago1
Zero-shot evaluation on other datasets, e.g., NYU, SUN, etc.
Closed 2 years ago2
The result of clip only
Closed 2 years ago2
About PCA for CLIP text embeddings
Closed 2 years ago2
Acquiring the updated scannetv2-labels.combined.tsv
Closed 2 years ago6
Test the result
Closed 2 years ago4
Tensors must be CUDA and dense
Closed 2 years ago4
Pre-training needs 3D annotations/supervision
Closed 2 years ago3
TypeError: max() received an invalid combination of arguments - got (out=NoneType, axis=NoneType, ), but expected one of:
Closed 2 years ago3