Jenna-Bai

Hello, Thanks so much for your work! I would like to use the dataset for my training. But I didn't find any annotations about the distortion types in the dataset....

Question about the training performance

4

Thanks so much for the release of code and dataset. I tried to retrain MVSSNet and Trufor on OpenSDI dataset under the IMDLBenco framework. After 3 epochs, the test_pixel-level Accuracy...

Question about HFDataset splits

3

In line 132 of data/hf_datasets.py, filtered_indices selects image paths with 'partial/fake' in the case of 'pixel' == True. Does this mean that only 'partial/fake' images are used for IMDL task?...

Thanks so much for the release of model weights! I noticed that there is a deepfake detector and a LLaVa model, which one produces the final prediction? I tried both...