Iuliia Tkachenko

ORCID: 0000-0001-5551-1105
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Digital Media Forensic Detection
  • Advanced Steganography and Watermarking Techniques
  • QR Code Applications and Technologies
  • Advanced Image and Video Retrieval Techniques
  • DNA and Biological Computing
  • Industrial Vision Systems and Defect Detection
  • Image Processing Techniques and Applications
  • Material Properties and Processing
  • Anomaly Detection Techniques and Applications
  • Advanced Malware Detection Techniques
  • Handwritten Text Recognition Techniques
  • Manufacturing Process and Optimization
  • Aesthetic Perception and Analysis
  • Image Processing and 3D Reconstruction
  • Generative Adversarial Networks and Image Synthesis
  • Advanced Image Processing Techniques
  • Digital Media and Visual Art
  • Image Retrieval and Classification Techniques
  • 3D Surveying and Cultural Heritage
  • Physical Education and Training Studies
  • Image and Object Detection Techniques
  • Nanofabrication and Lithography Techniques
  • Infrastructure Maintenance and Monitoring
  • Cellular Automata and Applications
  • Digital and Traditional Archives Management

Institut National des Sciences Appliquées de Lyon
2022-2025

Laboratoire d'Informatique en Images et Systèmes d'Information
2020-2025

Université Claude Bernard Lyon 1
2018-2025

Centre National de la Recherche Scientifique
2014-2025

Université Lumière Lyon 2
2020-2025

École Centrale de Lyon
2025

Laboratoire Hubert Curien
2018-2019

La Rochelle Université
2017

Laboratoire d'Informatique, de Robotique et de Microélectronique de Montpellier
2013-2016

Université de Montpellier
2013-2016

The quick response (QR) code was designed for storage information and high-speed reading applications. In this paper, we present a new rich QR that has two levels can be used document authentication. This code, named two-level public private levels. level is the same as standard level; therefore, it readable by any classical application. constructed replacing black modules specific textured patterns. It consists of encoded using q-ary with an error correction capacity. allows us not only to...

10.1109/tifs.2015.2506546 article EN IEEE Transactions on Information Forensics and Security 2015-12-09

The availability of high quality copy machines provides a large amount printed document counterfeits. Numerous authentication techniques based on security printing, graphical codes, hashing or local have been suggested earlier. In this paper, we propose novel system sensitivity new two level QR (2LQR) code to copying process. This 2LQR contains specific textured patterns, which are sensitive printing and processes. Therefore, it can be used detect unauthorized duplication. Experimental...

10.1109/icassp.2016.7472057 article EN 2016-03-01

Copy Detection Patterns (CDP) have received significant attention from academia and industry as a practical mean of detecting counterfeits. Their security level against sophisticated attacks has been studied theoretically practically in different research papers, but for reasons that will be explained below, the results are not fully conclusive. In addition, publicly available CDP datasets usable to evaluate performance authentication algorithms. short, apparently simple question: "are copy...

10.1109/wifs53200.2021.9648384 preprint EN 2021-12-07

Copy sensitive graphical codes are used as anti-counterfeiting solution in packaging and document protection. Their security is funded on a design hard-to-predict after print scan. In practice there exist different designs. Here random printed at the printer resolution considered. We suggest an estimation of such by using neural networks, in-trend approach which has however not been studied yet present context. this paper, we test state-of-the-art architecture efficient binarization...

10.1145/3335203.3335718 preprint EN 2019-07-02

The authentication of printed documents is a nowadays challenge. One the promising solutions for document use copy sensitive graphical codes that can offer data storage and support authentication. Both decoding physical are based on comparison between printed-and-scanned samples original numerical codes. In this paper we want to evaluate different existing correlation measures (Kendall Spearman) propose new Kendall weighted metric. We method by considering its ability decode stored messages...

10.2352/issn.2470-1173.2017.7.mwsf-335 article EN Electronic Imaging 2017-01-29

Verification of a document legitimacy is current important problem. In this paper we propose to use textured image containing visual message, which can be used for identification differences between printed legitimate and fake document. The suggested consists specific patterns should satisfy particular conditions in order give good recognition results after Print-and-Scan (P&S) process. possible by correlating the with either original or representative P&S process patterns. Several...

10.5281/zenodo.43810 article EN 2014-09-01

10.1016/j.jisa.2022.103133 article EN publisher-specific-oa Journal of Information Security and Applications 2022-03-11

Modern content-based image retrieval systems demonstrate rather good performance in identifying visually similar artworks. However, this task becomes more challenging when art history specialists aim to refine the list of artworks based on their criteria, thus we need train model reproduce refinement. In paper, propose an approach for improving paintings according specific simulated criteria. By approach, retrieve a request using ResNet50 and ANNOY algorithm. Then, simulate re-ranking two...

10.1109/iccvw60793.2023.00177 article EN 2023-10-02

The integrity check of printed and scanned documents is a hot topic these days. Several solutions were proposed for once. However forged quite often pass through double Print-and-Scan (P&S) process. P&S process impacts lot the shape color characters. Therefore, top OCR Engines cannot correctly recognize In this paper, we present problems that Tesseract Engine faces with when trying to characters twice. We suggest use PCA based character recognition method outperforms in our experiments. also...

10.1109/icdar.2017.392 preprint EN 2017-11-01

The falsification of hardcopy documents is a common problem these days. Numerous industrial and scientific solutions have been proposed to prevent falsifications. In this paper, we want study the security two level QR code which constructed using specific textured patterns that are sensitive print-and-scan impact. Such good candidate because it generalizes several concepts from codes. We take falsifier point view aims reconstruct fool authentication system detector. As contains sets same...

10.1109/wifs.2018.8630792 preprint EN 2018-12-01

The number of real world applications Quick Response (QR) codes is increasing, along with the extent stored information in QR codes. That why high density are commonly encountered daily life. Unfortunately, versions not very well readable by most smartphone code readers. main reading problems tilt correction, binarization and module recognition. Quite often cannot determine version due to correction problems. Binarization algorithms use global threshold methods, that sensitive print-and-scan...

10.1109/ipta.2014.7001950 article EN 2014-10-01

The high visual quality of modern deepfakes raises significant concerns about the trustworthiness digital media and makes facial tampering detection more challenging. Although current deep learning-based deepfake detectors achieve excellent results when tested on images or image sequences generated using known methods, generalization—where a trained model is tasked with detecting created previously unseen manipulation techniques—is still major challenge. In this paper, we investigate impact...

10.1109/icip49359.2023.10222043 article EN 2022 IEEE International Conference on Image Processing (ICIP) 2023-09-11

Recent papers point out the vulnerability of Copy Sensitive Graphical Codes (CSGC) while an opponent uses a neural network approach to estimate pattern then prints it as original one: such fake can successfully pass authentication test. Here, we show that GAN-like be even more powerful. A SRGAN-based architecture including superresolution tolerate lower scanner resolution and decode efficiently. Besides, use decoding technique perform test improve resistance CSGC estimation attacks.

10.1109/wifs47025.2019.9035104 preprint EN 2019-12-01

10.5220/0007565800002108 article EN Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications 2019-01-01

The authentication of printed documents is an important problem these days. Numerous techniques have been proposed in the relevant literature. One most promising solutions uses copy-sensitive graphical codes made particular patterns. two Level QR (2LQR) code specific textured patterns order to ensure sensitivity duplication process. test this based on comparing correlation values between original and printed-and-scanned with a pre-determined threshold. weakest feature technique rejection...

10.1109/icip.2018.8451839 preprint EN 2018-09-07

The number of medicine counterfeits increases significantly. This problem affects not only expensive medicines, but also some low cost ones. In this paper, we study the characteristics packages printed using rotogravure printing on blister foils and propose an authentication system that identifies equipment used for foils. process uses engraved cylinder a press. Each these elements has its own signature can be identification packaging authentication. Using constructed database, show impacts...

10.1109/wifs49906.2020.9360883 preprint EN 2020-12-06

Abstract A number of bridges have collapsed around the world over past years, with detrimental consequences on safety and traffic. To a large extend, such failures can be prevented by regular bridge inspections maintenance, tasks that fall in general category structural health monitoring (SHM). Those procedures are time labor consuming, which partly accounts for their neglect. Computer vision artificial intelligence (AI) methods potential to ease this burden, fully or partially automating...

10.1007/s43503-024-00041-7 article EN cc-by AI in Civil Engineering 2024-12-01

10.5220/0007565805770583 article EN cc-by-nc-nd Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications 2019-01-01
Coming Soon ...