Implementasi MobileNetV2 Pada Aplikasi Forensik Android Untuk Deteksi Citra AI-generated dengan Ketahanan Terhadap Transformasi Citra

Aryanahta Putra; Resmi Darni; Dony Novaliendry; Vikri Aulia

doi:10.47065/josh.v7i4.10666

Aryanahta Putra * Univesitas Negeri Padang, Padang, Indonesia
Resmi Darni Univesitas Negeri Padang, Padang, Indonesia
Dony Novaliendry Univesitas Negeri Padang, Padang, Indonesia
Vikri Aulia Univesitas Negeri Padang, Padang, Indonesia

(*) Corresponding Author

DOI: https://doi.org/10.47065/josh.v7i4.10666

Keywords: AI-Generate D Image; Digital Forensics; Mobile Application; MobileNetV2; Robustness Testing

Abstract

The development of Generative Artificial Intelligence has produced realistic synthetic images that are difficult to distinguish from authentic images through visual inspection. This study aims to implement an Android-based mobile digital forensics application for detecting AI-generate d images using MobileNetV2 converted to TensorFlow Lite for on-device inference. A quantitative-experimental approach used 2,000 images, consisting of 1,000 authentic and 1,000 AI-generate d images. The authentic class comprised 500 smartphone photographs and 500 GenImage samples, while the AI-generate d class included Stable Diffusion v1.4, Stable Diffusion v1.5, Wukong, and Midjourney images. The dataset was divided into 1,000 training, 500 validation, and 500 testing images. Training data were used for model development, validation data for model selection and threshold determination, and testing data only for final evaluation. Robustness testing used copies of 500 testing images without retraining. Under normal conditions with a threshold of 0.680, the model achieved 85.40% accuracy, 85.53% precision, 85.40% recall, and an 85.39% F1-score. JPEG q=65 compression produced 86.00% accuracy. The largest degradation occurred with 112 × 112 resizing combined with JPEG q=65, resulting in 62.40% accuracy, a decrease of 23.00 percentage points, and a 56.66% F1-score. The application performed local inference, displayed prediction labels and confidence scores, and stored detection history. This study contributes an on-device detection application and robustness evaluation using a consistent test subset, positioning the system as an initial detection aid rather than a final forensic verification tool.

Downloads

Download data is not yet available.

References

R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer, “High-Resolution Image Synthesis with Latent Diffusion Models,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022, pp. 10684–10695. doi: 10.1109/CVPR52688.2022.01042.

M. Zhu et al., “GenImage: A Million-Scale Benchmark for Detecting AI-generate d Image,” Adv. Neural Inf. Process. Syst., vol. 36, no. NeurIPS, 2023.

S. J. Nightingale and H. Farid, “AI-Synthesized Faces Are Indistinguishable from Real Faces and More Trustworthy,” Proc. Natl. Acad. Sci., vol. 119, no. 8, p. e2120481119, 2022, doi: 10.1073/pnas.2120481119.

J. J. Bird and A. Lotfi, “CIFAKE: Image Classification and Explainable Identification of AI-generate d Synthetic Images,” IEEE Access, vol. 12, pp. 26896–26909, 2024, doi: 10.1109/ACCESS.2024.3356122.

S. S. Baraheem and T. V Nguyen, “AI vs. AI: Can AI Detect AI-generate d Images?,” J. Imaging, vol. 9, no. 10, p. 199, 2023, doi: 10.3390/jimaging9100199.

U. Ojha, Y. Li, and Y. J. Lee, “Towards Universal Fake Image Detectors that Generalize Across Generative Models,” Proc. IEEE Comput. Soc. Conf. Comput. Vis. Pattern Recognit., vol. 2023-June, pp. 24480–24489, 2023, doi: 10.1109/CVPR52729.2023.02345.

R. Corvi, D. Cozzolino, G. Zingarini, G. Poggi, K. Nagano, and L. Verdoliva, “On the Detection of Synthetic Images Generate d by Diffusion Models,” 2023, IEEE. doi: 10.1109/ICASSP49357.2023.10095167.

Z. Wang et al., “DIRE for Diffusion-Generate d Image Detection,” pp. 22445–22455, 2023, [Online]. Available: https://openaccess.thecvf.com/content/ICCV2023/html/Wang_DIRE_for_Diffusion-Generate d_Image_Detection_ICCV_2023_paper.html

D. Cozzolino, G. Poggi, R. Corvi, M. Nießner, and L. Verdoliva, “Raising the Bar of AI-generate d Image Detection with CLIP,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2024, pp. 4356–4366. doi: 10.1109/CVPRW63382.2024.00439.

C. Tan, Y. Zhao, S. Wei, G. Gu, P. Liu, and Y. Wei, “Rethinking the Up-sampling Operations in CNN-Based Generative Network for Generalizable Deepfake Detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024, pp. 28130–28139. doi: 10.1109/CVPR52733.2024.02657.

Y. Li et al., “MaskSim: Detection of Synthetic Images by Masked Spectrum Similarity Analysis,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2024, pp. 3855–3865. doi: 10.1109/CVPRW63382.2024.00390.

J. Ricker, D. Lukovnikov, and A. Fischer, “AEROBLADE: Training-Free Detection of Latent Diffusion Images Using Autoencoder Reconstruction Error,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024, pp. 9130–9140. doi: 10.1109/CVPR52733.2024.00872.

H. Zhang, Q. He, X. Bi, W. Li, B. Liu, and B. Xiao, “Towards Universal AI-generate d Image Detection by Variational Information Bottleneck Network,” pp. 23828–23837, 2025, doi: 10.1109/cvpr52734.2025.02219.

F. Guillaro et al., “A Bias-Free Training Paradigm for More General AI-generate d Image Detection Real,” pp. 18685–18694.

Q. Bammey, “Synthbuster: Towards Detection of Diffusion Model Generate d Images,” IEEE Open J. Signal Process., vol. 5, pp. 1–9, 2024, doi: 10.1109/OJSP.2023.3337714.

C. Li et al., “RRDataset: Bridging the Gap Between Ideal and Real-world Evaluation: Benchmarking AI-generate d Image Detection in Challenging Scenarios,” pp. 20379–20389, 2025, [Online]. Available: http://arxiv.org/abs/2509.09172

D. Karageorgiou, S. Papadopoulos, I. Kompatsiaris, and E. Gavves, “Any-Resolution AI-generate d Image Detection by Spectral Learning,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025, pp. 18706–18717. doi: 10.1109/CVPR52734.2025.01743.

J. Park and A. Owens, “Community Forensics: Using Thousands of Generators to Train Fake Image Detectors,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025, pp. 8245–8257. doi: 10.1109/CVPR52734.2025.00772.

Z. Jia et al., “Secret Lies in Color: Enhancing AI-generate d Images Detection with Color Distribution Analysis,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025, pp. 13445–13454. doi: 10.1109/CVPR52734.2025.01255.

B. Chu, X. Xu, X. Wang, Y. Zhang, W. You, and L. Zhou, “FIRE: Robust Detection of Diffusion-Generate d Images via Frequency-Guided Reconstruction Error,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025, pp. 12830–12839. doi: 10.1109/CVPR52734.2025.01197.

H. Yu and B. Xu, “Multi-Modal Texture Fusion Network for Detecting AI-generate d Images,” Front. Artif. Intell., vol. 8, p. 1663292, 2025, doi: 10.3389/frai.2025.1663292.

M. Sandler, A. Howard, M. Zhu, A. Zhmoginov, and L. C. Chen, “MobileNetV2: Inverted Residuals and Linear Bottlenecks,” Proc. IEEE Comput. Soc. Conf. Comput. Vis. Pattern Recognit., pp. 4510–4520, 2018, doi: 10.1109/CVPR.2018.00474.

R. David et al., “TensorFlow Lite Micro: Embedded Machine Learning for TinyML Systems,” in Proceedings of Machine Learning and Systems, 2021, pp. 800–811. doi: 10.48550/arXiv.2010.08678.

Bila bermanfaat silahkan share artikel ini

Berikan Komentar Anda terhadap artikel Implementasi MobileNetV2 Pada Aplikasi Forensik Android Untuk Deteksi Citra AI-generated dengan Ketahanan Terhadap Transformasi Citra