Synthetic Data Generation for Emotional Depth Faces: Optimizing Conditional DCGANs via Genetic Algorithms in the Latent Space and Stabilizing Training with Knowledge Distillation
On the map: Knowledge distillation → General knowledge distillation methods
Abstract (arXiv)
Affective computing faces a major challenge: the lack of high-quality, diverse depth facial datasets for recognizing subtle emotional expressions. We propose a framework for synthetic depth face generation using an optimized GAN with Knowledge Distillation (EMA teacher models) to stabilize training, improve quality, and prevent mode collapse. We also apply Genetic Algorithms to evolve GAN latent vectors based on image statistics, boosting diversity and visual quality for target emotions. The approach outperforms GAN, VAE, GMM, and KDE in both diversity and quality. For classification, we extract and concatenate LBP, HOG, Sobel edge, and intensity histogram features, achieving 94% and 96% accuracy with XGBoost. Evaluation using FID, IS, SSIM, and PSNR shows consistent improvement over state-of-the-art methods.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in General knowledge distillation methods, one of 12 directions in Knowledge distillation. Builds on 20 of its 60 references that are in the map. First posted on arXiv 7 Aug 2025.
Builds on
20 of 60 references.
References · 20
- 2015Distilling the Knowledge in a Neural Networkcited by 10,324
- 2016XGBoost: A Scalable Tree Boosting Systemcited by 5,166
- 2016Improved Techniques for Training GANscited by 2,970
- 2014Conditional Generative Adversarial Netscited by 2,162
- 2024Comprehensive Exploration of Synthetic Data Generation: A Surveycited by 60
- 2024A survey of synthetic data augmentation methods in computer visioncited by 34
- 2023Example-based Motion Synthesis via Generative Motion Matchingcited by 16
- 20234D Facial Expression Diffusion Modelcited by 15
- 2019Kernel density estimation based sampling for imbalanced class distributioncited by 13
- 2017Photorealistic Facial Expression Synthesis by the Conditional Difference Adversarial Autoencodercited by 10
- 2020Joint Deep Learning of Facial Expression Synthesis and Recognitioncited by 6
- 2024SynFER: Towards Boosting Facial Expression Recognition with Synthetic Datacited by 6
8 more
- 20203D Dense Geometry-Guided Facial Expression Synthesis by Adversarial Learningcited by 5
- 2021CycleTransGAN-EVC: A CycleGAN-based Emotional Voice Conversion Model with Transformercited by 4
- 2025Synthetic Data Generation by Supervised Neural Gas Network for Physiological Emotion Recognition Datacited by 4
- 2020High-Accuracy Facial Depth Models derived from 3D Synthetic Datacited by 3
- 2021Altering Facial Expression Based on Textual Emotioncited by 2
- 2025Synthetic Data Generation of Body Motion Data by Neural Gas Network for Emotion Recognitioncited by 2
- 2021An augmentation strategy to mimic multi-scanner variability in MRIcited by 1
- 2023Victoria Amazonica Optimization (VAO): An Algorithm Inspired by the Giant Water Lily Plantcited by 1
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2026On Applicability of Synthetic Datasets for Facial Expression Recognition
- 2025Gen-AFFECT: Generation of Avatar Fine-grained Facial Expressions with Consistent identiTy
- 2024FlowVQTalker: High-Quality Emotional Talking Face Generation through Normalizing Flow and Quantization
- 2023GANDiffFace: Controllable Generation of Synthetic Datasets for Face Recognition with Realistic Variations
- 2024A Review of Human Emotion Synthesis Based on Generative Technology
- 2023Happy People -- Image Synthesis as Black-Box Optimization Problem in the Discrete Latent Space of Deep Generative Models
- 2025A Comparative Study on Synthetic Facial Data Generation Techniques for Face Recognition
- 2026Personalized Emotional Intelligence in Generative AI through Symbolic Affective Reasoning