Hi Authors, we're trying to reproduce the Hotels-8k results from the WACV 2024 paper. The download link in the README says that the link is removed. Could you re-upload the dataset, or point us to a current mirror?
For context: we have the raw FGVC8 Hotel-ID 2021 data from Kaggle, but HotelsDataset in datasets/hotels_8k.py expects
(a) per-hotel directory layout matching <hotel_id>/. and
(b) train.npy / val.npy / test.npy split files are in the data directory, neither of which is in the repo. If the broken link's tarball included these files, that would solve our reproduction.
Happy to mirror the data ourselves once we have it, if hosting is the issue.
Hi Authors, we're trying to reproduce the Hotels-8k results from the WACV 2024 paper. The download link in the README says that the link is removed. Could you re-upload the dataset, or point us to a current mirror?
For context: we have the raw FGVC8 Hotel-ID 2021 data from Kaggle, but HotelsDataset in datasets/hotels_8k.py expects
(a) per-hotel directory layout matching <hotel_id>/. and
(b) train.npy / val.npy / test.npy split files are in the data directory, neither of which is in the repo. If the broken link's tarball included these files, that would solve our reproduction.
Happy to mirror the data ourselves once we have it, if hosting is the issue.