Publications

@inproceedings{bondi2020training,

title = {Training Strategies and Data Augmentations in CNN-based DeepFake Video Detection},

author = {Luca Bondi and Edoardo Daniele Cannas and Paolo Bestagini and Stefano Tubaro},

doi = {10.1109/WIFS49906.2020.9360901},

booktitle = {IEEE International Workshop on Information Forensics and Security (WIFS)},

pages = {1-6},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{cannas2023enhancement,

title = {Enhancement Strategies For Copy-Paste Generation & Localization in RGB Satellite Imagery},

author = {Edoardo Daniele Cannas and Sriram Baireddy and Paolo Bestagini and Stefano Tubaro and Edward J. Delp},

doi = {10.1109/WIFS58808.2023.10374637},

booktitle = {IEEE International Workshop on Information Forensics and Security (WIFS)},

pages = {1-6},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{giganti2023multi,

title = {Multi-BVOC Super-Resolution Exploiting Compounds Inter-Connection},

author = {Antonio Giganti and Sara Mandelli and Paolo Bestagini and Marco Marcon and Stefano Tubaro},

doi = {10.23919/EUSIPCO58844.2023.10290089},

booktitle = {European Signal Processing Conference (EUSIPCO)},

pages = {1315-1319},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{bhagtani2023synthesized,

title = {Synthesized Speech Attribution Using The Patchout Spectrogram Attribution Transformer},

author = {Kratika Bhagtani and Emily R. Bartusiak and Amit Kumar Singh Yadav and Paolo Bestagini and Edward J. Delp},

url = {https://doi.org/10.1145/3577163.3595112},

doi = {10.1145/3577163.3595112},

isbn = {9798400700545},

booktitle = {ACM Workshop on Information Hiding and Multimedia Security (IH&MMSec)},

pages = {157–162},

publisher = {Association for Computing Machinery},

address = {Chicago, IL, USA},

series = {IH&MMSec '23},

abstract = {The malicious use of synthetic speech has increased with the recent availability of speech generation tools. It is important to determine whether a speech signal is authentic (spoken by a human) or is synthesized and to determine the generation method used to create it. Identifying the synthesis method is known as synthetic speech attribution. In this paper, we propose the use of a transformer deep learning method that analyzes mel-spectrograms for synthetic speech attribution. Our method known as Patchout Spectrogram Attribution Transformer (PSAT) can distinguish new, unseen speech generation methods from those seen during training. PSAT demonstrates high performance in attributing synthetic speech signals. Evaluation on the DARPA SemaFor Audio Attribution Dataset and the ASVSpoof2019 Dataset shows that our method achieves more than 95% accuracy in synthetic speech attribution and performs better than existing deep learning approaches.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{xiang2023mtn,

title = {MTN: Forensic Analysis of MP4 Video Files Using Graph Neural Networks},

author = {Ziyue Xiang and Amit Kumar Singh Yadav and Paolo Bestagini and Stefano Tubaro and Edward J. Delp},

booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshop},

pages = {963-972},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{giampiccolo2022timeb,

title = {A Time-Domain Virtual Bass Enhancement Circuital Model for Real-Time Music Applications},

author = {Riccardo Giampiccolo and Alberto Bernardini and Augusto Sarti},

doi = {10.1109/MMSP55362.2022.9949443},

booktitle = {IEEE 24th International Workshop on Multimedia Signal Processing},

address = {Shanghai, China},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{bondi2020trainingb,

title = {Training Strategies and Data Augmentations in CNN-based DeepFake Video Detection},

author = {Luca Bondi and Edoardo Daniele Cannas and Paolo Bestagini and Stefano Tubaro},

doi = {10.1109/WIFS49906.2020.9360901},

booktitle = {IEEE International Workshop on Information Forensics and Security (WIFS)},

pages = {1-6},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{cannas2023enhancementb,

title = {Enhancement Strategies For Copy-Paste Generation & Localization in RGB Satellite Imagery},

author = {Edoardo Daniele Cannas and Sriram Baireddy and Paolo Bestagini and Stefano Tubaro and Edward J. Delp},

doi = {10.1109/WIFS58808.2023.10374637},

booktitle = {IEEE International Workshop on Information Forensics and Security (WIFS)},

pages = {1-6},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{giganti2023multib,

title = {Multi-BVOC Super-Resolution Exploiting Compounds Inter-Connection},

author = {Antonio Giganti and Sara Mandelli and Paolo Bestagini and Marco Marcon and Stefano Tubaro},

doi = {10.23919/EUSIPCO58844.2023.10290089},

booktitle = {European Signal Processing Conference (EUSIPCO)},

pages = {1315-1319},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{bhagtani2023synthesizedb,

title = {Synthesized Speech Attribution Using The Patchout Spectrogram Attribution Transformer},

author = {Kratika Bhagtani and Emily R. Bartusiak and Amit Kumar Singh Yadav and Paolo Bestagini and Edward J. Delp},

url = {https://doi.org/10.1145/3577163.3595112},

doi = {10.1145/3577163.3595112},

isbn = {9798400700545},

booktitle = {ACM Workshop on Information Hiding and Multimedia Security (IH&MMSec)},

pages = {157–162},

publisher = {Association for Computing Machinery},

address = {Chicago, IL, USA},

series = {IH&MMSec '23},

abstract = {The malicious use of synthetic speech has increased with the recent availability of speech generation tools. It is important to determine whether a speech signal is authentic (spoken by a human) or is synthesized and to determine the generation method used to create it. Identifying the synthesis method is known as synthetic speech attribution. In this paper, we propose the use of a transformer deep learning method that analyzes mel-spectrograms for synthetic speech attribution. Our method known as Patchout Spectrogram Attribution Transformer (PSAT) can distinguish new, unseen speech generation methods from those seen during training. PSAT demonstrates high performance in attributing synthetic speech signals. Evaluation on the DARPA SemaFor Audio Attribution Dataset and the ASVSpoof2019 Dataset shows that our method achieves more than 95% accuracy in synthetic speech attribution and performs better than existing deep learning approaches.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{xiang2023mtnb,

title = {MTN: Forensic Analysis of MP4 Video Files Using Graph Neural Networks},

author = {Ziyue Xiang and Amit Kumar Singh Yadav and Paolo Bestagini and Stefano Tubaro and Edward J. Delp},

booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshop},

pages = {963-972},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}