Optical Flow Distillation: Towards Efficient and Stable Video Style Transfer

Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 12351)


Video style transfer techniques inspire many exciting applications on mobile devices. However, their efficiency and stability are still far from satisfactory. To boost the transfer stability across frames, optical flow is widely adopted, despite its high computational complexity, e.g. occupying over 97% inference time. This paper proposes to learn a lightweight video style transfer network via knowledge distillation paradigm. We adopt two teacher networks, one of which takes optical flow during inference while the other does not. The output difference between these two teacher networks highlights the improvements made by optical flow, which is then adopted to distill the target student network. Furthermore, a low-rank distillation loss is employed to stabilize the output of student network by mimicking the rank of input videos. Extensive experiments demonstrate that our student network without an optical flow module is still able to generate stable video and runs much faster than the teacher network.


Knowledge distillation Optical flow Video style transfer 



We thank anonymous reviewers for their helpful comments. Chang Xu was supported by the Australian Research Council under Project DE180101438.

Supplementary material (59.5 mb)
Supplementary material 1 (zip 60950 KB)


  1. 1.
    Butler, D.J., Wulff, J., Stanley, G.B., Black, M.J.: A naturalistic open source movie for optical flow evaluation. In: Fitzgibbon, A., Lazebnik, S., Perona, P., Sato, Y., Schmid, C. (eds.) ECCV 2012. LNCS, vol. 7577, pp. 611–625. Springer, Heidelberg (2012). Scholar
  2. 2.
    Chen, D., Liao, J., Yuan, L., Yu, N., Hua, G.: Coherent online video style transfer. In: ICCV, pp. 1105–1114 (2017)Google Scholar
  3. 3.
    Chen, D., Yuan, L., Liao, J., Yu, N., Hua, G.: Stylebank: an explicit representation for neural image style transfer. In: CVPR, pp. 1897–1906 (2017)Google Scholar
  4. 4.
    Chen, G., Choi, W., Yu, X., Han, T., Chandraker, M.: Learning efficient object detection models with knowledge distillation. In: NIPS, pp. 742–751 (2017)Google Scholar
  5. 5.
    Chen, H., Wang, Y., Xu, C., Xu, C., Tao, D.: Learning student networks via feature embedding. TNNLS (2020)Google Scholar
  6. 6.
    Chen, H., et al.: Data-free learning of student networks. In: ICCV (2019)Google Scholar
  7. 7.
    Dumoulin, V., Shlens, J., Kudlur, M.: A learned representation for artistic style. In: ICLR (2017)Google Scholar
  8. 8.
    Gao, C., Gu, D., Zhang, F., Yu, Y.: ReCoNet: real-time coherent video style transfer network. In: Jawahar, C.V., Li, H., Mori, G., Schindler, K. (eds.) ACCV 2018. LNCS, vol. 11366, pp. 637–653. Springer, Cham (2019). Scholar
  9. 9.
    Gao, W., Li, Y., Yin, Y., Yang, M.H.: Fast video multi-style transfer. In: WACV, pp. 3222–3230 (2020)Google Scholar
  10. 10.
    Gatys, L.A., Ecker, A.S., Bethge, M.: A neural algorithm of artistic style. arXiv preprint arXiv:1508.06576 (2015)
  11. 11.
    Gatys, L.A., Ecker, A.S., Bethge, M.: Image style transfer using convolutional neural networks. In: CVPR, pp. 2414–2423 (2016)Google Scholar
  12. 12.
    Gong, X., Chang, S., Jiang, Y., Wang, Z.: Autogan: neural architecture search for generative adversarial networks. In: ICCV, pp. 3224–3234 (2019)Google Scholar
  13. 13.
    Gupta, A., Johnson, J., Alahi, A., Fei-Fei, L.: Characterizing and improving stability in neural style transfer. In: ICCV, pp. 4067–4076 (2017)Google Scholar
  14. 14.
    He, T., Shen, C., Tian, Z., Gong, D., Sun, C., Yan, Y.: Knowledge adaptation for efficient semantic segmentation. In: CVPR, pp. 578–587 (2019)Google Scholar
  15. 15.
    He, Y., Zhang, X., Sun, J.: Channel pruning for accelerating very deep neural networks. In: ICCV (2017)Google Scholar
  16. 16.
    Heo, B., Kim, J., Yun, S., Park, H., Kwak, N., Choi, J.Y.: A comprehensive overhaul of feature distillation. In: ICCV, pp. 1921–1930 (2019)Google Scholar
  17. 17.
    Heo, B., Lee, M., Yun, S., Choi, J.Y.: Knowledge transfer via distillation of activation boundaries formed by hidden neurons. In: AAAI, vol. 33, pp. 3779–3787 (2019)Google Scholar
  18. 18.
    Hinton, G., Vinyals, O., Dean, J.: Distilling the knowledge in a neural network. arXiv preprint arXiv:1503.02531 (2015)
  19. 19.
    Huang, H., et al.: Real-time neural style transfer for videos. In: CVPR, pp. 783–791 (2017)Google Scholar
  20. 20.
    Huang, Z., Wang, N.: Like what you like: Knowledge distill via neuron selectivity transfer. arXiv preprint arXiv:1707.01219 (2017)
  21. 21.
    Jiao, J., Wei, Y., Jie, Z., Shi, H., Lau, R.W., Huang, T.S.: Geometry-aware distillation for indoor semantic segmentation. In: CVPR, pp. 2869–2878 (2019)Google Scholar
  22. 22.
    Johnson, J., Alahi, A., Fei-Fei, L.: Perceptual losses for real-time style transfer and super-resolution. In: Leibe, B., Matas, J., Sebe, N., Welling, M. (eds.) ECCV 2016. LNCS, vol. 9906, pp. 694–711. Springer, Cham (2016). Scholar
  23. 23.
    Li, Q., Jin, S., Yan, J.: Mimicking very efficient network for object detection. In: CVPR, pp. 6356–6364 (2017)Google Scholar
  24. 24.
    Li, Y., Fang, C., Yang, J., Wang, Z., Lu, X., Yang, M.H.: Universal style transfer via feature transforms. In: NIPS, pp. 386–396 (2017)Google Scholar
  25. 25.
    Liao, J., Yao, Y., Yuan, L., Hua, G., Kang, S.B.: Visual attribute transfer through deep image analogy. ACM Trans. Graph. (TOG) 36(4), 1–15 (2017)CrossRefGoogle Scholar
  26. 26.
    Liu, Y., Chen, K., Liu, C., Qin, Z., Luo, Z., Wang, J.: Structured knowledge distillation for semantic segmentation. In: CVPR, pp. 2604–2613 (2019)Google Scholar
  27. 27.
    Liu, Z., Wu, B., Luo, W., Yang, X., Liu, W., Cheng, K.-T.: Bi-real net: enhancing the performance of 1-Bit CNNs with improved representational capability and advanced training algorithm. In: Ferrari, V., Hebert, M., Sminchisescu, C., Weiss, Y. (eds.) ECCV 2018. LNCS, vol. 11219, pp. 747–763. Springer, Cham (2018). Scholar
  28. 28.
    Liu, Z., Sun, M., Zhou, T., Huang, G., Darrell, T.: Rethinking the value of network pruning. In: ICLR (2019)Google Scholar
  29. 29.
    Lu, M., Zhao, H., Yao, A., Chen, Y., Xu, F., Zhang, L.: A closed-form solution to universal style transfer. In: ICCV, October 2019Google Scholar
  30. 30.
    Lu, M., Zhao, H., Yao, A., Xu, F., Chen, Y., Zhang, L.: Decoder network over lightweight reconstructed feature for fast semantic style transfer. In: ICCV, pp. 2469–2477 (2017)Google Scholar
  31. 31.
    Marszałek, M., Laptev, I., Schmid, C.: Actions in context. In: CVPR, pp. 2929–2936. IEEE Computer Society (2009)Google Scholar
  32. 32.
    Romero, A., Ballas, N., Kahou, S.E., Chassang, A., Gatta, C., Bengio, Y.: Fitnets: hints for thin deep nets. In: ICLR (2015)Google Scholar
  33. 33.
    Ruder, M., Dosovitskiy, A., Brox, T.: Artistic style transfer for videos. In: Rosenhahn, B., Andres, B. (eds.) GCPR 2016. LNCS, vol. 9796, pp. 26–36. Springer, Cham (2016). Scholar
  34. 34.
    Ruder, M., Dosovitskiy, A., Brox, T.: Artistic style transfer for videos and spherical images. IJCV 126(11), 1199–1219 (2018)MathSciNetCrossRefGoogle Scholar
  35. 35.
    Saputra, M.R.U., de Gusmao, P.P., Almalioglu, Y., Markham, A., Trigoni, N.: Distilling knowledge from a deep pose regressor network. In: ICCV (2019)Google Scholar
  36. 36.
    Shen, M., Han, K., Xu, C., Wang, Y.: Searching for accurate binary neural architectures. In: ICCV Neural Architects Workshop (2019)Google Scholar
  37. 37.
    Shu, H., et al.: Co-evolutionary compression for unpaired image translation. In: ICCV, pp. 3235–3244 (2019)Google Scholar
  38. 38.
    Wang, C., Kong, C., Lucey, S.: Distill knowledge from NRSfM for weakly supervised 3D pose learning. In: ICCV, pp. 743–752 (2019)Google Scholar
  39. 39.
    Wang, W., Xu, J., Zhang, L., Wang, Y., Liu, J.: Consistent video style transfer via compound regularization. In: AAAI, pp. 12233–12240 (2020)Google Scholar
  40. 40.
    Wang, Y., Xu, C., Xu, C., Tao, D.: Adversarial learning of portable student networks. In: AAAI (2018)Google Scholar
  41. 41.
    Wei, Y., Pan, X., Qin, H., Ouyang, W., Yan, J.: Quantization mimic: towards very tiny CNN for object detection. In: Ferrari, V., Hebert, M., Sminchisescu, C., Weiss, Y. (eds.) ECCV 2018. LNCS, vol. 11212, pp. 274–290. Springer, Cham (2018). Scholar
  42. 42.
    Yang, Z., et al.: Cars: continuous evolution for efficient neural architecture search. In: CVPR, pp. 1829–1838 (2020)Google Scholar
  43. 43.
    Yim, J., Joo, D., Bae, J., Kim, J.: A gift from knowledge distillation: fast optimization, network minimization and transfer learning. In: CVPR, pp. 4133–4141 (2017)Google Scholar
  44. 44.
    Zagoruyko, S., Komodakis, N.: Paying more attention to attention: improving the performance of convolutional neural networks via attention transfer. In: ICLR (2017)Google Scholar
  45. 45.
    Zhou, S., Wu, Y., Ni, Z., Zhou, X., Wen, H., Zou, Y.: DoReFa-Net: Training low bitwidth convolutional neural networks with low bitwidth gradients. arXiv preprint arXiv:1606.06160 (2016)

Copyright information

© Springer Nature Switzerland AG 2020

Authors and Affiliations

  1. 1.Noah’s Ark LabHuawei TechnologiesShenzhenChina
  2. 2.School of Computer Science, Faculty of EngineeringUniversity of SydneySydneyAustralia

Personalised recommendations