# A Brief Review for Compression and Transfer Learning Techniques in DeepFake Detection

arXiv: https://arxiv.org/abs/2504.21066 · HTML: https://mlsanity.com/paper/2504.21066/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 12 of its 26 references that are in the map.
- First posted on arXiv 29 Apr 2025.

## Builds on (most cited first, 12 of 12)

- [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://mlsanity.com/paper/1409.1556/) (2014)
- [CNN-generated images are surprisingly easy to spot... for now](https://mlsanity.com/paper/1912.11035/) (2019)
- [VL-Adapter: Parameter-Efficient Transfer Learning for Vision-and-Language Tasks](https://mlsanity.com/paper/2112.06825/) (2021)
- [Pruning and Quantization for Deep Neural Network Acceleration: A Survey](https://mlsanity.com/paper/2101.09671/) (2021)
- [The Creation and Detection of Deepfakes: A Survey](https://mlsanity.com/paper/2004.11138/) (2020)
- [Media Forensics and DeepFakes: an overview](https://mlsanity.com/paper/2001.06564/) (2020)
- [SpotTune: Transfer Learning through Adaptive Fine-tuning](https://mlsanity.com/paper/1811.08737/) (2018)
- [Deep Learning for Deepfakes Creation and Detection: A Survey](https://mlsanity.com/paper/1909.11573/) (2019)
- [Domain-adaptive deep network compression](https://mlsanity.com/paper/1709.01041/) (2017)
- [Towards Optimal Trade-offs in Knowledge Distillation for CNNs and Vision Transformers at the Edge](https://mlsanity.com/paper/2407.12808/) (2024)
- [From deepfake to deep useful: risks and opportunities through a systematic literature review](https://mlsanity.com/paper/2311.15809/) (2023)
- [Reducing Inference Energy Consumption Using Dual Complementary CNNs](https://mlsanity.com/paper/2412.01039/) (2024)

## Built on by (most cited first, 1 of 1)

- [Frugal Machine Learning for Energy-efficient, and Resource-aware Artificial Intelligence](https://mlsanity.com/paper/2506.01869/) (2025)

## Abstract (arXiv)

Training and deploying deepfake detection models on edge devices offers the advantage of maintaining data privacy and confidentiality by processing it close to its source. However, this approach is constrained by the limited computational and memory resources available at the edge. To address this challenge, we explore compression techniques to reduce computational demands and inference time, alongside transfer learning methods to minimize training overhead. Using the Synthbuster, RAISE, and ForenSynths datasets, we evaluate the effectiveness of pruning, knowledge distillation (KD), quantization, fine-tuning, and adapter-based techniques. Our experimental results demonstrate that both compression and transfer learning can be effectively achieved, even with a high compression level of 90%, remaining at the same performance level when the training and validation data originate from the same DeepFake model. However, when the testing dataset is generated by DeepFake models not present in the training set, a domain generalization issue becomes evident.