Learning Rich Features for Image Manipulation Detection

  title={Learning Rich Features for Image Manipulation Detection},
  author={Peng Zhou and Xintong Han and Vlad I. Morariu and Larry S. Davis},
  journal={2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  • Peng ZhouXintong Han L. Davis
  • Published 13 May 2018
  • Computer Science
  • 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition
Image manipulation detection is different from traditional semantic object detection because it pays more attention to tampering artifacts than to image content, which suggests that richer features need to be learned. [] Key Method The other is a noise stream that leverages the noise features extracted from a steganalysis rich model filter layer to discover the noise inconsistency between authentic and tampered regions.

Towards Effective Image Manipulation Detection with Proposal Contrastive Learning

This work proposes Proposal Contrastive Learning (PCL), a two-stream architecture by extracting two types of global features from RGB and noise views respectively that can reduce manual labeling costs and promote more generalizable features in image manipulation detection.

Multi-Modality Image Manipulation Detection

A novel two-stream multi-modality image manipulation detection model (MM-net) that abandons the way of fusion to exploit one stream to guide the learning of the other stream through attention mechanism and builds the first tampered dataset of natural images based on GANs.

Digital image manipulation detection with weak feature stream

The performance of the Faster R-CNN network with weak feature stream has been improved significantly in F1 score and localization of the tampering region and a weak feature extraction network based on multi-scale residual network is established.

Holistic Image Manipulation Detection using Pixel Co-occurrence Matrices

A novel approach to holistically detect tampered images using a combination of pixel co-occurrence matrices and deep learning, which is agnostic to the type of manipulation and classifies an image as tampered or untampered.

Image Tamper Detection Based on Two-Stream Attention Faster R-CNN

An end-to-end image tamper detection model called Two-Stream Attention Net (TSA-Net) based on RGB-N and Faster R-CNN, where the Region of Interest (RoI) pooling inRGB-N is replaced with RoI align and the detection ability of the model in small targets is improved.

Local Pixel Manipulation Detection with Deep Neural Networks

This thesis presents an approach based on deep learning that significantly improves the possibilities to detect manipulations, and not only classifies the images, for being manipulated or not, but also localizes where the manipulations occur.

Image Forgery Detection Using Tamper-Guided Dual Self-Attention Network with Multiresolution Hybrid Feature

An end-to-end fully convolutional neural network with multiresolution hybrid features from RGB stream and noise stream is proposed, which can simultaneously effectively achieve pixel- level forgery localization and image-level forgery detection while maintaining higher detection accuracy and stronger robustness.

Auto-Focus Contrastive Learning for Image Manipulation Detection

After learning the AF-CL network by minimizing the distance between the representations of corresponding views, the learned network is able to automatically focus on the manipulated region and its surroundings and sufficiently explore their trace relations for accurate manipulation detection.

Constrained R-Cnn: A General Image Manipulation Detection Model

This work proposes a coarse-to-fine architecture named Constrained R-CNN for complete and accurate image forensics that effectively discriminates manipulated regions for the next manipulation classification and coarse localization.

A Skip Connection Architecture for Localization of Image Manipulations

An encoder-decoder based network where representations from early layers in the encoder are fused by skip pooling with representations of the last layer of the decoder to use for manipulation detection and can achieve a significantly better performance than the state-of-the-art methods and baselines.



Exploiting Spatial Structure for Localizing Manipulated Image Regions

A high confidence detection framework which can localize manipulated regions in an image by learning the boundary discrepancy between manipulated and non-manipulated regions with the combination of LSTM and convolution layers.

A deep learning approach to detection of splicing and copy-move forgeries in images

  • Y. RaoJ. Ni
  • Computer Science
    2016 IEEE International Workshop on Information Forensics and Security (WIFS)
  • 2016
A new image forgery detection method based on deep learning technique, which utilizes a convolutional neural network to automatically learn hierarchical representations from the input RGB color images to outperforms some state-of-the-art methods.

A Deep Learning Approach to Universal Image Manipulation Detection Using a New Convolutional Layer

A universal forensic approach to performing manipulation detection using deep learning that can automatically learn how to detect multiple image manipulations without relying on pre-selected features or any preprocessing is proposed.

Image Region Forgery Detection: A Deep Learning Approach

A two stage deep learning approach to learn features in order to detect tampered images in different image formats using a Stacked Autoencoder model to learn the complex feature for each individual patch so that the detection can be conducted more accurately.

R-FCN: Object Detection via Region-based Fully Convolutional Networks

This work presents region-based, fully convolutional networks for accurate and efficient object detection, and proposes position-sensitive score maps to address a dilemma between translation-invariance in image classification and translation-variance in object detection.

Median Filtering Forensics Based on Convolutional Neural Networks

This work proposes a median filtering detection method based on convolutional neural networks (CNNs), which can automatically learn and obtain features directly from the image and achieves significant performance improvements, especially in the cut-and-paste forgery detection.

Bilinear CNN Models for Fine-Grained Visual Recognition

We propose bilinear models, a recognition architecture that consists of two feature extractors whose outputs are multiplied using outer product at each location of the image and pooled to obtain an

Deep Regionlets for Object Detection

This paper proposes a "region selection network" and a "gating network" that serves as a guidance on where to select regions to learn the features from in Regionlet and achieves comparable state-of-the-art results.

Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks

This work introduces a Region Proposal Network (RPN) that shares full-image convolutional features with the detection network, thus enabling nearly cost-free region proposals and further merge RPN and Fast R-CNN into a single network by sharing their convolutionAL features.

Rich Models for Steganalysis of Digital Images

A novel general strategy for building steganography detectors for digital images by assembling a rich model of the noise component as a union of many diverse submodels formed by joint distributions of neighboring samples from quantized image noise residuals obtained using linear and nonlinear high-pass filters.