Summary of Compressed Deepfake Video Detection Based on 3d Spatiotemporal Trajectories, by Zongmei Chen et al.
Compressed Deepfake Video Detection Based on 3D Spatiotemporal Trajectories
by Zongmei Chen, Xin Liao, Xiaoshuai Wu, Yanxiang Chen
First submitted to arxiv on: 28 Apr 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI); Multimedia (cs.MM)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed deepfake video detection method, based on 3D spatiotemporal trajectories, aims to address the limitations of existing methods that primarily focus on uncompressed videos. The approach utilizes a robust 3D model to construct features from both 2D and 3D frames, mitigating the influence of head rotation angles or lighting conditions. A sequential analysis method based on phase space motion trajectories is designed to explore feature differences between genuine and fake faces in deepfake videos. Extensive experiments validate the performance of the proposed method on compressed deepfake benchmarks, verifying the robustness of well-designed features. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper proposes a new way to detect deepfakes in videos that have been compressed. Most current methods only work well with uncompressed videos, but this one is designed to be more reliable and accurate even when the video has been heavily compressed. The method uses special 3D models to analyze both the shape of faces and the movement of heads to tell real from fake. It’s tested on several different datasets and shows promising results. |
Keywords
» Artificial intelligence » Spatiotemporal
