Pornographic Image Recognition via Weighted Multiple Instance Learning

Abstract

In the era of Internet, recognizing pornographic images is of great significance for protecting children's physical and mental health. However, this task is very challenging as the key pornographic contents (e.g., breast and private part) in an image often lie in local regions of small size. In this paper, we model each image as a bag of regions, and follow a multiple instance learning (MIL) approach to train a generic region-based recognition model. Specifically, we take into account the regions' degree of pornography, and make three main contributions. First, we show that based on very few annotations of the key pornographic contents in a training image, we can generate a bag of properly sized regions, among which the potential positive regions usually contain useful contexts that can aid recognition. Second, we present a simple quantitative measure of a region's degree of pornography, which can be used to weigh the importance of different regions in a positive image. Third, we formulate the recognition task as a weighted MIL problem under the convolutional neural network framework, with a bag probability function introduced to combine the importance of different regions. Experiments on our newly collected large scale dataset demonstrate the effectiveness of the proposed method, achieving an accuracy with 97.52% true positive rate at 1% false positive rate, tested on 100K pornographic images and 100K normal images.

The application of neural networks in the work of forensic experts in child abuse cases

“…The model showed the accuracy of almost 98%. Jin et al [25] are the authors of the model that reached over 97% accuracy in the classification of pornographic and non-pornographic images. As mentioned before, the COPINE Scale is used to categorize pornographic materials, which makes it a useful tool that can provide data on the character of pornographic materials involving minors and its impact on the recipient by inducing sexual stimulation.…”

“…Jin i wsp. [25] są autorami modelu osiągającego ponad 97% skuteczności klasyfikacji zdjęć pornograficznych i niepornograficznych. Jak zaprezentowano wcześniej, skala COPINE jest używana do kategoryzacji materiałów pornograficznych, przez co jest użytecznym narzędziem mogącym dostarczyć przesłanek o charakterze pornografii z udziałem małoletnich oraz jej wpływie na odbiorcę poprzez wywołanie pobudzenia seksualnego.…”

Automated Nudity Recognition using Very Deep Residual Learning Network

“…The field of pornography content recognition has not been an exception and several attempts have been made to enhance the previous results by the aid of DL like those reported in [3], [4], [6], [10], [25].…”

“…Later in [25] a deep learning-based approach was proposed as a decision support tool to aid evaluation of sexual assaults. In [4], the authors proposed a Weighted Multiple Instance Learning (WMIL) scheme to be fused with a CNN model for the purpose of detecting regions containing pornography contents.…”

“…Three different evaluation metrics were used to assess the performance of the proposed binary classification models (nudity vs. non-nudity) as Accuracy, AUC, F-score formulated in (1), (2), (3), and (4).…”