Video Frame Interpolation Using Deep Convolutional Neural Network

  • Varghese Mathai
  • Arun Baby
  • Akhila Sabu
  • Jeexson Jose
  • Bineeth KuriakoseEmail author
Conference paper
Part of the Lecture Notes in Computational Vision and Biomechanics book series (LNCVB, volume 30)


Video frame interpolation fuse several low-resolution (LR) frames into one high-resolution (HR) frame. The existing methods for video frame interpolation use optical flow method to determine motion in a scene, but computation using optical flow method is difficult, which can lead to artifacts in the output video. In many applications where we use video footages, there is a similarity in the content of footages. This similarity in content recommends that using some kind of context-aware approach can do better interpolation than the different existing interpolation techniques. We propose such a context-aware approach for video interpolation, the video frame interpolation using convolutional neural networks. In this proposed method, neighboring images are given as input to an end-to-end convolutional neural network which interpolates a frame between them. A comparative analysis of video interpolation technique using proposed RGB model and HSV model using metric standards such as SSIM, PSNR, and MSE is also included in the proposed method.


Video frame interpolation Convolutional neural network Dataset Loss function 


  1. 1.
    Kang K, Ouyang W, Li H, Wang X (2016) Object detection from video tubelets with convolutional neural networks. In: IEEE conference on computer vision and pattern recognition (CVPR), pp 817–825 Google Scholar
  2. 2.
    Dong C, Loy CC, He K, Tang X (2016) Image super-resolution using deep convolutional networks. IEEE Trans Pattern Anal Mach Intell 38(2):295307CrossRefGoogle Scholar
  3. 3.
    Hung K-W, Siu W-C (2011) Fast video interpolation/upsampling using linear motion model. In: 18th IEEE international conference on image processingGoogle Scholar
  4. 4.
    Guo D, Lu Z (2016) Motion-compensated Frame Interpolation with weighted motion estimation and hierarchical vector refinement. Neurocomputing 181CrossRefGoogle Scholar
  5. 5.
    Ghutke RC, Naveen C, Satpute VR (2016) Temporal video frame interpolation using new cubic motion compensation technique. In: IEEE international conference on signal processing and communications (SPCOM)Google Scholar
  6. 6.
    Saravanan G, Yamuna G, Nandhini S (2016) Real time implementation of RGB to HSV/HSI/HSL and its reverse color space models. In: IEEE international conference on communication and signal processing, 6–8 April 2016Google Scholar
  7. 7.
    Kingma DP, Ba J (2014) Adam: a method for stochastic optimization. CoRR abs/1412.6980 available at
  8. 8.
    Ronneberger O, Fischer P, Brox T (2015) U-Net: convolutional networks for biomedical image segmentation, CoRR abs/1505.04597, available at
  9. 9.
    Song BC, Jeong SC, Choi Y (2011) Video super-resolution algorithm using bi-directional overlapped block motion compensation and on-the-fly dictionary training. IEEE Trans Circuits Syst Video Technol 21(3):274–285CrossRefGoogle Scholar
  10. 10.
    Kotevski Z, Mitrevski P (2009) Experimental comparison of PSNR and SSIM metrics for video quality estimation. In: Davcev D, Gmez JM (eds) ICT innovations. Springer, Berlin, HeidelbergGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  • Varghese Mathai
    • 1
  • Arun Baby
    • 1
  • Akhila Sabu
    • 1
  • Jeexson Jose
    • 1
  • Bineeth Kuriakose
    • 1
    Email author
  1. 1.Department of Computer Science and EngineeringMITSErnakulamIndia

Personalised recommendations