Deep Learning Features Fusion with Classical Image Features for Image Access

Depending on the society, the access to the adult content can create social problems. This paper thus proposes a fusion approach for image based adult content filtering. The proposed approach merges the Deep Learning (DL) architecture and classical hand crafted feature extraction approaches. From the DL, we fuse the rich feature extraction capabilities of the Convolutional Neural Networks (CNNs) with the Correlograms features. We optimize the classification by integrating and modifying the Correlograms into skin Correlograms. The results show an increased performance by combining the DL learnt features with the classical hand crafted features. From an evaluation, the proposed approach achieves an Accuracy of 0.93. This work thus motivates the usage of classical hand crafted features to be exploited in the DL architectures for segmentation and detection scenarios.

@article{Khan2018,
  title     = {Deep Learning Features Fusion with Classical Image Features for Image Access},
  journal   = {International Journal of Advanced Computer Science and Applications},
  volume    = {9},
  number    = {7},
  year      = {2018},
  publisher = {The Science and Information Organization},
  author    = {Rehan Ullah Khan},
  doi       = {10.14569/IJACSA.2018.090707},
  url       = {https://doi.org/10.14569/IJACSA.2018.090707}
}