Applying depthwise separable and multi-channel convolutional neural networks of varied kernel size on semantic trajectories

  • Antonios KaratzoglouEmail author
  • Nikolai Schnell
  • Michael Beigl
Brain inspired Computing&Machine Learning Applied Research-BISMLARE


Convolutional neural networks (CNN) have become due to their outstanding performance in the past few years rapidly the standard approach when it comes to processing 2D data as these can be found in the image recognition and classification domain. Recent research shows that CNN models can handle 1D data, such as temporal sequences (e.g., speech and text), with a similar high performance as well. This fact motivated our present idea to apply convolutional networks for modeling human semantic trajectories and predicting future locations. Our work consists of three parts. The first part evaluates the performance of a standard spatial CNN in comparison with a vanilla feed-forward, a recurrent and a long short-term memory network (LSTM) at two different semantic representation levels. In the second part, we explore in depth the impact of the kernel size and propose a multi-channel convolutional approach based on kernels of varied size. Finally, part three investigates the depthwise factorization of the convolutional layer with regard to training time and test accuracy. Altogether, it can be shown that convolutional networks are able to outperform the competition, with the channel number as well as the kernel size being the most significant hyperparameters.


Convolutional neural networks Depthwise separable convolution Multi-channel convolution Semantic trajectories Location prediction 


Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.


  1. 1.
    Alvares LO, Bogorny V, Kuijpers B, de Macedo JAF, Moelans B, Vaisman A (2007) A model for enriching trajectories with semantic geographical information. In: Proceedings of the 15th annual ACM international symposium on advances in geographic information systems. ACM, New York, p 22Google Scholar
  2. 2.
    Arriaga O, Valdenegro-Toro M, Plöger P (2017) Real-time convolutional neural networks for emotion and gender classification. CoRR arXiv:1710.07557
  3. 3.
    Ashbrook D, Starner T (2003) Using GPS to learn significant locations and predict movement across multiple users. Pers Ubiquitous Comput 7(5):275–286. CrossRefGoogle Scholar
  4. 4.
    Bogorny V, Renso C, Aquino AR, Lucca Siqueira F, Alvares LO (2014) Constant—a conceptual data model for semantic trajectories of moving objects. Trans GIS 18(1):66–88CrossRefGoogle Scholar
  5. 5.
    Cao X, Cong G, Jensen CS (2010) Mining significant semantic locations from GPS data. Proc VLDB Endow 3(1–2):1009–1020. CrossRefGoogle Scholar
  6. 6.
    Cart (2018) Site planning and revenue prediction: optimizing food truck locations in New York City (online). Accessed 29 July 2019
  7. 7.
    Chollet F (2017) Xception: deep learning with depthwise separable convolutions. Preprint arXiv:1610.02357
  8. 8.
    Chollet F (2017) Xception: deep learning with depthwise separable convolutions. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 1251–1258Google Scholar
  9. 9.
    Collobert R, Weston J, Bottou L, Karlen M, Kavukcuoglu K, Kuksa P (2011) Natural language processing (almost) from scratch. J Mach Learn Res 12(Aug):2493–2537zbMATHGoogle Scholar
  10. 10.
    dos Santos C, Gatti M (2014) Deep convolutional neural networks for sentiment analysis of short texts. In: Proceedings of COLING 2014, the 25th international conference on computational linguistics: technical papers, pp 69–78Google Scholar
  11. 11.
    Eagle N, Pentland AS (2006) Reality mining: sensing complex social systems. Pers Ubiquit Comput 10(4):255–268CrossRefGoogle Scholar
  12. 12.
    Elragal A, El-Gendy N (2013) Trajectory data mining: integrating semantics. J Enterp Inf Manag 26(5):516–535. CrossRefGoogle Scholar
  13. 13.
    Etter V, Kafsi M, Kazemi E (2012) Been there, done that: what your mobility traces reveal about your behavior. In: Proceedings of mobile data challenge by nokia workshop. 10th PerComGoogle Scholar
  14. 14.
    Facebook (2018) Offline trajectories. United States Patent Application 20,180,352,383Google Scholar
  15. 15.
    Fan RC, Yang X, Fay JD (2003) Using location data to determine traffic information. US Patent 6,594,576Google Scholar
  16. 16.
    Gao Q, Zhou F, Zhang K, Trajcevski G, Luo X, Zhang F (2017) Identifying human mobility via trajectory embeddings. In: Proceedings of the 26th international joint conference on artificial intelligence. AAAI Press, New York, pp 1689–1695Google Scholar
  17. 17.
    Google (2016) Do i stay or do i go now? Google maps has the answer in one tap (online). Accessed 29 July 2019
  18. 18.
    Graves A, Mohamed A, Hinton G (2013) Speech recognition with deep recurrent neural networks. In: 2013 IEEE international conference on acoustics, speech and signal processing. IEEE, New York, pp 6645–6649Google Scholar
  19. 19.
    Gutfreund H, Mezard M (1988) Processing of temporal sequences in neural networks. Phys Rev Lett 61(2):235MathSciNetCrossRefGoogle Scholar
  20. 20.
    Kaiser L, Gomez AN, Chollet F (2017) Depthwise separable convolutions for neural machine translation. Preprint arXiv:1706.03059
  21. 21.
    Kalchbrenner N, Grefenstette E, Blunsom P (2014) A convolutional neural network for modelling sentences. Preprint arXiv:1404.2188
  22. 22.
    Karatzoglou A (2019) Evolutionary optimization on artificial neural networks for predicting the user’s future semantic location. In: Macintyre J, Iliadis L, Maglogiannis I, Jayne C (eds) Engineering applications of neural networks. Springer, Cham, pp 379–390CrossRefGoogle Scholar
  23. 23.
    Karatzoglou A, Jablonski A, Beigl M (2018) A Seq2Seq learning approach for modeling semantic trajectories and predicting the next location. In: Proceedings of the 26th ACM SIGSPATIAL international conference on advances in geographic information systems, SIGSPATIAL’18. ACM, New York, pp 528–531.
  24. 24.
    Karatzoglou A, Köhler D, Beigl M (2018) Semantic-enhanced multi-dimensional Markov chains on semantic trajectories for predicting future locations. Sensors 18(10):3582. CrossRefGoogle Scholar
  25. 25.
    Karatzoglou A, Lamp SC, Beigl M (2017) Matrix factorization on semantic trajectories for predicting future semantic locations. In: 2017 IEEE 13th international conference on wireless and mobile computing, networking and communications (WiMob), pp 1–7.
  26. 26.
    Karatzoglou A, Schnell N, Beigl M (2018) A convolutional neural network approach for modeling semantic trajectories and predicting future locations. In: Kůrková V, Manolopoulos Y, Hammer B, Iliadis L, Maglogiannis I (eds) Artificial neural networks and machine learning—ICANN 2018. Springer, Cham, pp 61–72CrossRefGoogle Scholar
  27. 27.
    Karatzoglou A, Sentürk H, Jablonski A, Beigl M (2017) Applying artificial neural networks on two-layer semantic trajectories for predicting the next semantic location. In: International conference on artificial neural networks. Springer, New York, pp 233–241CrossRefGoogle Scholar
  28. 28.
    Kim Y (2014) Convolutional neural networks for sentence classification. Preprint arXiv:1408.5882
  29. 29.
    LeCun Y, Bengio Y et al (1995) Convolutional networks for images, speech, and time series. Handb Brain Theory Neural Netw 3361(10):1995Google Scholar
  30. 30.
    Lv J, Li Q, Wang X (2016) T-conv: a convolutional neural network for multi-scale taxi trajectory prediction. Preprint arXiv:1611.07635
  31. 31.
    Mathworks (2018) Convolutional neural network (online). Accessed 19 Feb 2018
  32. 32.
    Newsroom TE (2017) A new business intelligence emerges: (online). Accessed 29 July 2019
  33. 33.
    Quercia D, Lathia N, Calabrese F, Di Lorenzo G, Crowcroft J (2010) Recommending social events from mobile phone location data. In: 2010 IEEE international conference on data mining, pp 971–976.
  34. 34.
    Ratti C, Frenchman D, Pulselli RM, Williams S (2006) Mobile landscapes: using location data from cell phones for urban analysis. Environ Plan 33(5):727–748. CrossRefGoogle Scholar
  35. 35.
    Sifre L, Mallat S (2013) Rotation, scaling and deformation invariant scattering for texture discrimination. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 1233–1240Google Scholar
  36. 36.
    Skyhook (2019) How IoT can benefit from location (online). Accessed 29 July 2019
  37. 37.
    Song X, Kanasugi H, Shibasaki R (2016) Deeptransport: prediction and simulation of human mobility and transportation mode at a citywide level. In: Proceedings of 25th international joint conference on artificial intelligence, pp 2618–2624Google Scholar
  38. 38.
    Spaccapietra S, Parent C, Damiani ML, de Macêdo JAF, Porto F, Vangenot C (2008) A conceptual view on trajectories. Data Knowl Eng 65(1):126–146. CrossRefGoogle Scholar
  39. 39.
    Uberbacher EC, Mural RJ (1991) Locating protein-coding regions in human dna sequences by a multiple sensor-neural network approach. Proc Natl Acad Sci 88(24):11261–11265CrossRefGoogle Scholar
  40. 40.
    Ying JJC, Lee WC, Tseng VS (2014) Mining geographic–temporal–semantic patterns in trajectories for location prediction. ACM Trans Intell Syst Technol 5(1):2:1–2:33. CrossRefGoogle Scholar
  41. 41.
    Ying JJC, Lee WC, Weng TC, Tseng VS (2011) Semantic trajectory mining for location prediction. In: Proceedings of the 19th ACM SIGSPATIAL international conference on advances in geographic information systems. ACM, New York, pp 34–43Google Scholar

Copyright information

© Springer-Verlag London Ltd., part of Springer Nature 2019

Authors and Affiliations

  1. 1.Robert Bosch GmbH, Chassis Systems Control, Advance EngineeringAbstattGermany
  2. 2.Karlsruhe Institute of TechnologyKarlsruheGermany

Personalised recommendations