# Extracting deep local features to detect manipulated images of human faces

arXiv: https://arxiv.org/abs/1911.13269 · HTML: https://mlsanity.com/paper/1911.13269/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 29 Nov 2019.

## Built on by (most cited first, 8 of 8)

- [Countering Malicious DeepFakes: Survey, Battleground, and Horizon](https://mlsanity.com/paper/2103.00218/) (2021)
- [Towards Solving the DeepFake Problem : An Analysis on Improving DeepFake Detection using Dynamic Face Augmentation](https://mlsanity.com/paper/2102.09603/) (2021)
- [Towards Generalizable and Robust Face Manipulation Detection via Bag-of-local-feature](https://mlsanity.com/paper/2103.07915/) (2021)
- [Generalized Facial Manipulation Detection with Edge Region Feature Extraction](https://mlsanity.com/paper/2102.01381/) (2021)
- [WWW: Where, Which and Whatever Enhancing Interpretability in Multimodal Deepfake Detection](https://mlsanity.com/paper/2408.02954/) (2024)
- [DeepFakes: Detecting Forged and Synthetic Media Content Using Machine Learning](https://mlsanity.com/paper/2109.02874/) (2021)
- [Deepfake Detection for Facial Images with Facemasks](https://mlsanity.com/paper/2202.11359/) (2022)
- [Datasets, Clues and State-of-the-Arts for Multimedia Forensics: An Extensive Review](https://mlsanity.com/paper/2401.06999/) (2024)

## Abstract (arXiv)

Recent developments in computer vision and machine learning have made it possible to create realistic manipulated videos of human faces, raising the issue of ensuring adequate protection against the malevolent effects unlocked by such capabilities. In this paper we propose local image features that are shared across manipulated regions are the key element for the automatic detection of manipulated face images. We also design a lightweight architecture with the correct structural biases for extracting such features and derive a multitask training scheme that consistently outperforms image class supervision alone. The trained networks achieve state-of-the-art results in the FaceForensics++ dataset using significantly reduced number of parameters and are shown to work well in detecting fully generated face images.