Augmented convolutional feature maps for robust CNN-based camera model identification

Identifying the model of the camera that captured an image is an important forensic problem. While several algorithms have been proposed to accomplish this, their performance degrades significantly if the image is subject to post-processing. This is problematic since social media applications and ph...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:2017 IEEE International Conference on Image Processing (ICIP) S. 4098 - 4102
Hauptverfasser: Bayar, Belhassen, Stamm, Matthew C.
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 01.09.2017
Schlagworte:
Adaptation models
Camera identification
Cameras
constrained convolutional layer
convolutional neural networks
deep convolutional features
Feature extraction
Forensics
Robustness
Training
Transform coding
ISSN:2381-8549
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Identifying the model of the camera that captured an image is an important forensic problem. While several algorithms have been proposed to accomplish this, their performance degrades significantly if the image is subject to post-processing. This is problematic since social media applications and photo-sharing websites typically resize and recompress images. In this paper, we propose a new convolutional neural network based approach to performing camera model identification that is robust to resampling and recompression. To accomplish this, we propose a new approach to low-level feature extraction that uses both a constrained convolutional layer and a nonlinear residual feature extractor in parallel. The feature maps produced by both of these layers are then concatenated and passed to subsequent convolutional layers for further feature extraction. Experimental results show that our proposed approach can significantly improve camera model identification performance in resampled and recompressed images.
ISSN:2381-8549
DOI:10.1109/ICIP.2017.8297053