Advertisement

Novel Scene Recognition Using TrainDetector

  • Sebastien MambouEmail author
  • Ondrej Krejcar
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11401)

Abstract

Our ability to process the image keeps improving day by day, since the introduction of deep learning. Lastly, this contributed to the advance of object recognition through a Convolutional neural network and Place recognition, which is our concern in this paper. Through this research, it was observed a complexity in the extraction of the correct and relevant features for scene recognition. To address this issue, we extracted at the pixel level several subareas which contain more color intensity than other parts, and we went through each image once to build the feature representation of it. We also noticed that several available models based on Convolution Neural Network requires a Graphics Processing Units (GPU) for their implementation and are difficult to train. We propose in this paper, a novel Scene Recognition method using Single-Shot-Detector (SSD), Multi-modal Local-Receptive-Field (MM-LRF) and Extreme-Learning-Machine (ELM) that we named TrainDetector. It outperforms the state-of-the-art techniques when we apply it to three well-known scene recognition Datasets.

Keywords

SSD MM-LRF ELM TrainDetector 

Notes

Acknowledgement

The work and the contribution were supported by the SPEV project “Smart Solutions in Ubiquitous Computing Environments 2018”, University of Hradec Kralove, Faculty of Informatics and Management, Czech Republic.

References

  1. 1.
    Luo, J., Boutell, M.: Natural scene classification using overcomplete ICA. Pattern Recognit. 38(10), 1507–1519 (2005)CrossRefGoogle Scholar
  2. 2.
    Mundhenk, T., Flores, A., Hoffman, H.: Classification and segmentation of orbital space based objects against terrestrial distractors for the purpose of finding holes in Shape from Motion 3D reconstruction. In: Proceedings of SPIE, vol. 9025 (2014)Google Scholar
  3. 3.
    Wang, Q., Chen, L., Shen, D.: Group-wise registration of large image dataset by hierarchical clustering and alignment. In: Proceedings of SPIE, vol. 7259, no. 35 (2009)Google Scholar
  4. 4.
    Newsam, S., Kamath, C.: Comparing shape and texture features for pattern recognition in simulation data. Electron. Imaging 5672, 106–117 (2005)Google Scholar
  5. 5.
    Kunter, M., Knorr, S., Krutz, A., Sikora, T.: Unsupervised object segmentation for 2D to 3D conversion. In: Proceedings of SPIE, vol. 7237 (2009)Google Scholar
  6. 6.
    Yu, K., Lin, Y., Lafferty, J.: Learning image representations from the pixel level via hierarchical sparse coding. http://dblp.uni-trier.de/db/conf/cvpr/cvpr2011.html. Accessed 2011
  7. 7.
    Asif, U., Bennamoun, M., Sohel, F.: Efficient RGB-D object categorization using cascaded ensembles of randomized decision trees. http://dblp.uni-trier.de/db/conf/icra/icra2015.html. Accessed 2015
  8. 8.
    Huang, G.-B., Zhu, Q.-Y., Siew, C.-K.: Extreme learning machine: a new learning scheme of feedforward neural networks. http://ieeexplore.ieee.org/document/1380068. Accessed 2004
  9. 9.
    Lazebnik, S., Schmid, C., Ponce, J.: 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition - Volume 2, CVPR 2006 (2006)Google Scholar
  10. 10.
    Quattoni, A., Torralba, A.: Recognizing indoor scenes. http://people.csail.mit.edu/torralba/publications/indoor.pdf. Accessed 2009
  11. 11.
    Xiao, J., Hays, J., Ehinger, K., Oliva, A., Torralba, A.: SUN database: Large-scale scene recognition from abbey to zoo. http://ieeexplore.ieee.org/document/5539970. Accessed 2010
  12. 12.
    Huang, G.-B., Bai, Z., Kasun, L., Vong, C.: Local receptive fields based extreme learning machine. IEEE Comput. Intell. Mag. 10(2), 18–29 (2015)CrossRefGoogle Scholar
  13. 13.
    Preim, B., Botha, C.: Image analysis for medical visualization. https://sciencedirect.com/science/article/pii/b9780124158733000043. Accessed 2014
  14. 14.
    Razavian, A., Azizpour, H., Sullivan, J., Carlsson, S.: CNN features off-the-shelf: an astounding baseline for recognition. In: Computer Vision and Pattern Recognition, pp. 512–519 (2014)Google Scholar
  15. 15.
    Zhou, B., Lapedriza, À., Xiao, J., Torralba, A., Oliva, A.: learning deep features for scene recognition using places database. https://papers.nips.cc/paper/5349-learning-deep-features-for-scene-recognition-using-places-database. Accessed 2014
  16. 16.
    Cimpoi, M., Maji, S., Vedaldi, A.: Deep filter banks for texture recognition and segmentation. http://dblp.uni-trier.de/db/conf/cvpr/cvpr2015.html. Accessed 2015
  17. 17.
    Zuo, Z., Wang, G., Shuai, B., Zhao, L., Yang, Q., Jiang, X.: Learning discriminative and shareable features for scene classification. https://link.springer.com/chapter/10.1007/978-3-319-10590-1_36. Accessed 2014
  18. 18.
    Xie, G.-S., Zhang, X.-Y., Yan, S., Liu, C.-L.: Hybrid CNN and dictionary-based models for scene recognition and domain adaptation. IEEE Trans. Circuits Syst. Video Technol. 27, 1263–1274 (2016)CrossRefGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  1. 1.Center for Basic and Applied Research, Faculty of Informatics and ManagementUniversity of Hradec KraloveHradec KraloveCzech Republic

Personalised recommendations