davide-coccomini / Combining-EfficientNet-and-Vision-Transformers-for-Video-Deepfake-Detection

Code for Video Deepfake Detection model from "Combining EfficientNet and Vision Transformers for Video Deepfake Detection" presented at ICIAP 2021.
https://dl.acm.org/doi/abs/10.1007/978-3-031-06433-3_19
MIT License
237 stars 60 forks source link

Inquiry Regarding Dataset Architecture and Organization #50

Closed 747135205 closed 8 months ago

747135205 commented 11 months ago

I hope this message finds you well. I would like to express my gratitude for the significant contributions you have made to the project. Your work has been truly inspiring and has provided me with valuable insights into the tasks I am aiming to accomplish.

Unfortunately, due to various reasons, I have encountered challenges in accessing the dataset hosted on dfdc.ai. Instead, I have been able to locate the dfdc dataset within the Kaggle project. I am reaching out to kindly inquire if you could provide some insights into the specific architecture of the dataset and the roles played by each of its components.

Additionally, I have a specific question regarding the contents of the DFDC folder. Are the data stored in this folder a combination of video frames with both FAKE and REAL attributes? Or do all REAL video frames exclusively reside in the ORIGINAL folder, while the remaining FAKE video frames are distributed across folders such as DFDC, FACE++, and others?

I understand that your time is valuable, and I genuinely appreciate your consideration in addressing my inquiries amidst your busy schedule.

Thank you for your time and assistance.

davide-coccomini commented 10 months ago

Thank you for your appreciation, we are glad to know that our work was useful! Unfortunately, you are not the first to be having problems downloading the full dataset from the DFDC site and we have no control over this nor can we publish it as it would violate the authors' rules. From Kaggle I believe there is the complete training set but the test set with the labels is missing, which you could take from the dfdc.ai site.

Regarding the structure of the dataset we worked with, in the Original folder, there are all the pristine FF++ videos. In the DFDC folder, there are both pristine and fake videos of DFDC. All other folders contain FF++'s fake videos. Following the extract_crops process, each folder contains images representing the faces detected by detect_faces in each frame of each video.

In case you need further clarification, please ask, I will try to answer you as soon as possible!

Sun-researcher commented 8 months ago

屏幕截图 2024-03-22 172733

Sun-researcher commented 8 months ago

屏幕截图 2024-03-22 173445

Sun-researcher commented 8 months ago

I followed the solution you suggested in #18, but encountered this problem, but I did not change the program