This paper studies how to build deep-fake detectors that both generalize to unseen image generators and provide transparent explanations of their decisions, and proposes a hybrid CLIP-Diffusion architecture that combines a frozen CLIP back-bone with a classifier operating on intermediate features and a diffusion-based reconstruction error head.
This work investigates what cues are exploited by foundation-model-based detectors to distinguish real images from diffusion-generated ones and suggests that foundation-model-based detectors succeed by capturing non-semantic low-to-mid frequency distributional discrepancies between real and diffusion-generated images.
Advances in the realism of synthetic media created with generative adversarial networks (GANs), diffusion models, and face manipulation tools has created an increased demand for well-established deepfake detection systems that can detect many different types of manipulation artifacts. However, most single model deepfake detectors are not very robust because they rely on specific forensic cues and do not adapt well to shifts in how synthesis occurs. We present DeepFakeBuster as a confidence-calibrated adaptive ensemble for deepfake image detection by fusing together heterogeneous deep learning models built around detecting complementary forensic cues e.g., spatial inconsistencies, boundary artifacts, noise residuals, semantic consistency, and frequency-domain features. In contrast to traditional ensemble approaches that use static averaging of detector outputs, our proposed framework utilizes reliability aware adaptive fusion where the contribution of each detector to the fused output is adjusted dynamically through the use of reliability priors derived from validation and input-specific confidence estimates. Our experimental evaluation on a dataset comprised of 192,000 authentic and manipulated images shows that our ensemble significantly outperforms both individual constituent detectors as well as static fusion baselines, with an overall accuracy of 97.8% for the evaluated conditions. Additionally, an interpretable forensic analysis module provides visual and quantitative indicators associated with manipulation-sensitive regions. The findings suggest that confidence-aware heterogeneous ensemble learning represents a promising direction for robust deepfake detection.
Rachana Patil, R. Shinde, S. Patil et al.· Scientific Reports· 0 citations
A data-generation pipeline that captions real photographs with a vision–language model and regenerates them with modern text-to-image systems, producing semantically aligned real/synthetic pairs that isolate generative artifacts from image content is described.
GFRE employs a lightweight autoencoder to model the reconstructability of image representations, producing a reconstruction signal that is inherently generator-agnostic and transferable across diverse generative processes, enabling efficient and scalable deployment.
Qinghui He, Haifeng Zhang, ∗. BoLiu et al.· 0 citations
Deep learning-based generative models have made a major leap forward in the world of image generation with the help
of Artificial Intelligence. One of the most notable of these developments is text-to-image synthesis, which can
automatically generate images based on natural language descriptions. In this work, an AI-based image generation system
is introduced that utilizes a Stable Diffusion model fine-tuned with Low-Rank Adaptation (LoRA) for domain-specific
image generation. The main idea of the proposed system is to combine the text encoding of CLIP, the latent compression
of Variational Autoencoder (VAE), and the denoising ability of diffusion to create images that are both semantically
relevant and visually coherent based on text prompts. The proposed approach was tested on a Pokemon image-caption
dataset for fine-tuning the pre-trained Stable Diffusion model and its effectiveness evaluated. The study shows that the
diffusion-based architectures outperform the traditional GAN based methods in terms of image quality, training stability,
semantic alignment, and output diversity. The main advantage of LoRA fine-tuning was the substantial decrease in
computational load, which involved updating just a small fraction of trainable parameters without compromising the
model's performance. Experimental results indicated that successful images of Pokemon could be generated, and that the
images were consistent with the text attributes such as color, type, and appearance. The results demonstrate that SD+LoRA
is an efficient and scalable domain-specific text-to-image generation system. The research underscores the rising
significance of diffusion-based generative AI in digital content creation, imaginative design, entertainment, and cleverness
in visual generation systems
Ankam Pavitra, R. Mallikharjun, Dr. L Jagadeesh Naik· International Journal of Dru...· 0 citations
Deepfakes pose growing risks to information integrity, yet many detectors perform well only on the datasets they were trained on and remain opaque to human analysts. A robust, explainable detection framework is presented that combines a CNN backbone for extracting spatial artifacts with an LSTM module for modeling temporal inconsistencies across frames. To make decisions auditable, the architecture incorporates Grad-CAM for spatial heatmaps, SHAP for quantitative feature attribution, and LIME for local surrogate explanations. The system was trained primarily on FaceForensics++ with stratified sampling and augmentation to reduce dataset bias and evaluated on multiple external benchmarks to assess cross-domain generalization. Experimental results show strong detection metrics, such as accuracy of 96.3%, precision of 95.8%, recall of 96.7%, and an F1-score of 96.2%, along with robust performance under JPEG compression, Gaussian noise, and FGSM adversarial attacks. By coupling high detection accuracy with transparent explanations, the proposed approach enhances forensic decision support and increases practical readiness for content verification systems.
Lastone Banda, Esther J.· International Journal of Dat...· 0 citations
