A5028899632- Digital Media Forensic Detection
- Advanced Steganography and Watermarking Techniques
- Advanced Image and Video Retrieval Techniques
- Chaos-based Image/Signal Encryption
- Image Retrieval and Classification Techniques
- Video Analysis and Summarization
- Face recognition and analysis
- Generative Adversarial Networks and Image Synthesis
- Advanced Data Compression Techniques
- Biometric Identification and Security
- Digital and Cyber Forensics
- Visual Attention and Saliency Detection
- Image Processing Techniques and Applications
- Anomaly Detection Techniques and Applications
- Text and Document Classification Technologies
- Computer Graphics and Visualization Techniques
- Multimodal Machine Learning Applications
- Misinformation and Its Impacts
- Image and Video Quality Assessment
- Mobile Ad Hoc Networks
- Law in Society and Culture
- Benford’s Law and Fraud Detection
- Advanced Image Processing Techniques
- Advanced Malware Detection Techniques
- Music and Audio Processing
University of Trento
2016-2025
Consorzio Nazionale Interuniversitario per le Telecomunicazioni
2018-2024
Universität Innsbruck
2020
Digital forensics is a relatively new research area which aims at authenticating digital media by detecting possible forgeries. Indeed, the ever increasing availability of multimedia data on web, coupled with great advances reached computer graphical tools, makes modification an image and creation visually compelling forgeries easy task for any user. This in turns creates need reliable tools to validate trustworthiness represented information. In such context, we present here RAISE, large...
Current developments in computer vision and deep learning allow to automatically generate hyper-realistic images, hardly distinguishable from real ones. In particular, human face generation achieved a stunning level of realism, opening new opportunities for the creative industry but, at same time, scary scenarios where such content can be maliciously misused. Therefore, it is essential develop innovative methodologies tell apart generated multimedia, possibly able follow evolution continuous...
Abstract The dependability of visual information on the web and authenticity digital media appearing virally in social platforms has been raising unprecedented concerns. As a result, last years multimedia forensics research community pursued ambition to scale forensic analysis real-world web-based open systems. This survey aims at describing work done so far shared data, covering three main aspects: techniques performing source identification integrity verification uploaded networks,...
Deep fakes became extremely popular in the last years, also thanks to their increasing realism. Therefore, there is need measures human's ability distinguish between real and synthetic face images when confronted with cutting-edge creation technologies. We describe design results of a perceptual experiment we have conducted, where wide diverse group volunteers has been exposed produced by state-of-the-art Generative Adversarial Networks (namely, PG-GAN, StyleGAN, StyleGAN2). The outcomes...
We describe a new forensic technique for distinguishing between computer generated and human faces in video. This identifies tiny fluctuations the appearance of face that result from changes blood flow. Because these pulse, they are unlikely to be found imagery. use absence or presence this physiological signal distinguish faces.
Morphological operators are widely used in binary image processing for several purposes, such as removing noise, detecting contours or particular structures, and regularizing shapes. In particular, morphological filters largely adopted scanned documents to correct the artifacts caused by acquisition binarization, well other processing. this paper, we propose a novel approach forensics detection of filtering on images. The proposed technique exploits some mathematical properties two basic...
The analysis of JPEG compressed images is one the most studied problems in image forensics, because extensive use and characteristic traces left by such coding operation. In this paper, we propose a novel statistical framework for identification previous multiple aligned compressions estimation quality factors applied. method has been tested on different datasets forensic scenarios, where up to three are considered. Moreover, both case double triple encoding with factors, compression history...
Mathematical morphology provides a large set of powerful non-linear image operators, widely used for feature extraction, noise removal or enhancement. Although morphological filters might be to remove artifacts produced by manipulations, both on binary and gray level documents, little effort has been spent towards their forensic identification. In this paper we propose non-trivial extension deterministic approach originally detecting erosion dilation images. The proposed operates grayscale...
The amount of multimedia content shared everyday, combined with the level realism reached by recent fake-generating technologies, threatens to impair trustworthiness online information sources. process uploading and sharing data tends hinder standard media forensic analyses, since multiple re-sharing steps progressively hide traces past manipulations. At same time though, new are introduced platforms themselves, enabling reconstruction history digital objects, possible applications in flow...
The rise of AI-generated synthetic media, or deepfakes, has introduced unprecedented opportunities and challenges across various fields, including entertainment, cybersecurity, digital communication. Using advanced frameworks such as Generative Adversarial Networks (GANs) Diffusion Models (DMs), deepfakes are capable producing highly realistic yet fabricated content, while these advancements enable creative innovative applications, they also pose severe ethical, social, security risks due to...
Providing reliable and efficient routing in presence of relative movement motivates the introduction awareness to improve performance existing position-based schemes vehicular ad-hoc networks. The proposed algorithm represents a modification well-known GPSR which exploits information about order next forwarding node decision. Performance evaluation protocol underlines promising robust basis for designing strategy suitable automotive scenario.
Source identification for digital content is one of the main branches image forensics. It relies on extraction photo-response non-uniformity (PRNU) noise as a unique intrinsic fingerprint that efficiently characterizes device which generated content. Such estimated difference between and its de-noised version obtained via denoising filter processing. This paper proposes performance comparison different filters source purposes. In particular, results achieved with sophisticated 3D are...
Intrinsic statistical properties of natural uncompressed images are used in image forensics for detecting the traces previous processing operations. In this paper, we propose novel forensic detectors JPEG compression stored formats, based on a theoretical analysis Benford-Fourier coefficients computed 8 × block-Discrete Cosine Transform (DCT) domain. fact, distribution such is derived theoretically both under hypotheses no and with certain quality factor, allowing computation respective...
In this paper, we propose a novel approach for face spoofing detection using the high-order Local Derivative Pattern from Three Orthogonal Planes (LDP-TOP). The proposed method is not only simple to derive and implement, but also highly efficient, since it takes into account both spatial temporal information in different directions of subtle movements. According experimental results, outperforms state-of-the-art methods on three reference datasets, namely Idiap REPLAY-ATTACK, CASIA-FASD, MSU...
In this letter, the design of a miniaturized monopolar prefractal antenna for 3.4-3.6 GHz WiMax band is presented. The geometrical configuration antenna, printed on planar dielectric substrate, has been synthesized by means particle swarm optimizer in order to minimize linear dimensions device and obtain voltage standing wave ratio (VSWR) values within specifications. results numerical simulations are shown compared, terms VSWR, with experimental measurements
The manipulation of text on a sign or billboard is relatively easy to do in way that perceptually convincing. When planar surface and imaged under perspective projection, the undergoes specific distortion. manipulated, it unlikely precisely satisfy this geometric mapping. We describe technique for detecting if an image obeys expected deviations from which are used as evidence tampering.
The widespread of social networking services allows users to share and quickly spread an enormous amount digital contents. Currently, a low level security trustworthiness is applied such information, whose reliability cannot be taken for granted due the large availability image editing software which allow any user easily manipulate This has huge impact on deception users, opinion can seriously influenced by altered media. In this work, we face challenge verifying online news analyzing...
In this paper, we propose a forensic technique that is able to detect the application of median filter 1D data. The method relies on deterministic mathematical properties filter, which lead identification specific relationships among sample values cannot be found in filtered sequences. Hence, their presence analyzed sequence allows excluding filter. Owing its nature, ensures 0% false negatives, and although positives (sequences not classified as filtered) are theoretically possible,...
A novel method for watermarking and ciphering color images, based on the joint use of a key-dependent wavelet transform with secure cryptographic scheme, is presented. The system allows to watermark encrypted data without requiring knowledge original also cipher watermarked damaging embedded signal. Since different areas proposed domain are used encryption watermarking, extraction hidden information can be performed deciphering cover it possible decipher removing watermark. Experimental...