Enhancing Privacy in Computer Vision Applications: An Emotion Preserving Approach to Obfuscate Faces

Nejad, Bijan Shahbaz; Roch, Peter; Handte, Marcus; Marrón, Pedro José

doi:10.1007/978-3-031-20716-7_7

Cited by 2 publications

(2 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The authors reported emotion preservation accuracy using CIAGAN and Deep Privacy was 52% and 60% respectively. Shahbaz et al [17] employed a image processing method for emotion-preserving de-identification. The authors used a proxy set of faces generated using StyleGAN and later used those proxy faces to obfuscate the target face.…”

Section: Literature Reviewmentioning

confidence: 99%

See 1 more Smart Citation

Emotion-Aware Face De-identification with Generative Adversarial Networks

Shopon,

Gavrilova

2024

Preprint

View full text Add to dashboard Cite

This paper presents an innovative end-to-end emotion-preserving de-identification approach utilizing a Generative Adversarial Network (GAN), employing the StyleGAN architecture. The method produces natural-looking de-identified images by generating a synthetic face image dataset and utilizing the DeepFace model for gender classification and representative image selection. An enhanced SimSwap framework enhances emotion preservation, accompanied by a novel loss function tailored to emotional expression preservation. The deep face model serves to classify and recognize emotional expressions in both original and de-identified swapped images. Explicit preservation of emotional expressions during face-swapping is achieved through attribute preservation loss minimization. The paper conducts a detailed ablation study to demonstrate the superiority of the proposed GAN components. The method exhibits superior performance in accuracy and emotion preservation compared to recent face de-identification methods.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

“…The results show that the proposed attribute preservation loss performs extremely well. We also have compared our method with the state-of-the-art recent facial de-identification system EPIC [17] (see Table 4). Both methods were evaluated using the LightFace pre-trained model.…”

Section: Experimentation and Analysismentioning

confidence: 99%