VLN-CE icon indicating copy to clipboard operation
VLN-CE copied to clipboard

Vision-and-Language Navigation in Continuous Environments using Habitat

Results 25 VLN-CE issues
Sort by recently updated
recently updated
newest added

Hi, I'm getting this error when training seq2seq baseline: ``` miniconda3/envs/vlnce-py3.6/lib/python3.6/site-packages/gym/spaces/discrete.py", line 36, in __init__ assert n > 0, "n (counts) have to be positive" AssertionError: n (counts) have to...

I have downloaded the preprocessed R2R datasets from [this official website](https://jacobkrantz.github.io/vlnce/data). In {split}_gt.json.gz, the field 'actions' contains ground truth actions, which should produce the coordinates stored in the field 'locations'....

Hello, I am running the statement 'python download_map. py -- task habitat - o data/scene_datasets/mp3d/-- id 17DRP5sb8fy'. After downloading, there is an additional `tasks` folder, but there is only one...

Is there any code related to the back-translation data augmentation method in this project? The paper mentions the use of this method, but I don't seem to see any related...

Great work! I want to follow it! But I am confused about the {split}_gt.json when I look into the action_sequence. First, you don't provide the lookup table about the action...