Research on Sensitive Information Recognition Algorithm Based on Deep Neural Network

Shuaina Huang; Karpovich Dmitry Semyonovich

doi:10.62051/ijcsit.v8n1.04

Authors

Shuaina Huang
Karpovich Dmitry Semyonovich

DOI:

https://doi.org/10.62051/ijcsit.v8n1.04

Keywords:

Sensitivity information recognition, Feature fusion, Deep neural network

Abstract

The spread of pornographic, violent, and politically sensitive images on social networks poses serious risks to youth well-being and social stability, making accurate detection essential for public safety. However, existing models often compromise representational capacity to meet computational constraints, while challenges such as illumination variation, changes in sensitive-region scale, and background interference further limit recognition accuracy. To address these issues, this study proposes a lightweight sensitive-image recognition framework. A high-order convolutional module is introduced to extract fine-grained features and suppress irrelevant background noise. A feature-guided fusion module is further designed to integrate texture and deep semantic features, improving robustness against lighting fluctuations and noise. The overall architecture builds on a compact high-resolution network enhanced with depthwise separable convolutions, MBConv layers, and Ghost Modules to significantly reduce parameters while maintaining strong performance. Experiments on a multi-class sensitive-image dataset verify the model’s efficiency and effectiveness.

Downloads

Download data is not yet available.

References

[1] Mishra D, Panda S. A Comparative Analysis of Pornography Detection Models to Prevent Gender Violence [M]//Communication technology and gender violence. Cham: Springer International Publishing, 2023: 99-107.

[2] Garcia M B, Revano T F, Habal B G M, et al. A pornographic image and video filtering application using optimized nudity recognition and detection algorithm [C]//2018 IEEE 10th international conference on humanoid, nanotechnology, information technology, communication and control, environment and management (HNICEM). IEEE, 2018: 1-5.

[3] jiao F, Gao W, Duan L, et al. Detecting adult image using multiple features [C]//2001 International conferences on info-tech and info-net. proceedings (Cat. No. 01EX479). IEEE, 2001: 378-383.

[4] Jones M J, Rehg J M. Statistical color models with application to skin detection [J]. International journal of computer vision, 2002, 46(1): 81-96.

[5] Jedynak B, Zheng H, Daoudi M. Statistical models for skin detection [C]//2003 Conference on computer vision and pattern recognition workshop. IEEE, 2003: 92-92.

[6] Lee J S, Kuo Y M, Chung P C, et al. Naked image detection based on adaptive and extensible skin color model [J]. Pattern recognition, 2007, 40(8): 2261-2270.

[7] Chen L C, Papandreou G, Kokkinos I, et al. Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected crfs [J]. IEEE transactions on pattern analysis and machine intelligence, 2017, 40(4): 834-848.

[8] Liu H, Guo R Y. Detection and identification of SAWH pipe weld defects based on X-ray image and CNN [J]. Chinese journal of scientific instrument, 2018, 39(4): 247-256.

[9] Wang Y H, Wang J, Tan X. Pornographic image recognition by strongly-supervised deep multiple instance learning [C]//2016 IEEE international conference on image processing (ICIP). IEEE, 2016: 4418-4422.

[10] Moustafa M. Applying deep learning to classify pornographic images and videos [J]. arxiv preprint arxiv:1511.08899, 2015.

[11] Ou X, Ling H, Yu H, et al. Adult image and video recognition by a deep multicontext network and fine-to-coarse strategy [J]. ACM transactions on intelligent systems and technology (TIST), 2017, 8(5): 1-25.

[12] Xiang T Z, Xia G S, Bai X, et al. Image stitching by line-guided local warping with global similarity constraint [J]. Pattern recognit, 2018, 77: 113-125.

[13] Cheng F, Wang S L, Wang X Z, et al. A global and local context integration DCNN for adult image classification [J]. Pattern recognition, 2019, 96: 106983.

[14] Wang X, Cheng F, Wang S, et al. Adult image classification by a local-context aware network [C]//2018 25th IEEE international conference on image processing (ICIP). IEEE, 2018: 2989-2993.