Filippos Kokkinos

Filippos Kokkinos

Ph.D. Student

University College London


I am currently a Ph.D. student at UCL under the supervision of Prof. Iasonas Kokkinos. Before that, I was a member of the Computational Imaging Group (CIG) at Skoltech, where I collaborated with Prof. Stamatis Lefkimmiatis on learnable regularizers. My research interests include machine learning, 3D reconstruction, computational photography, and inverse imaging problems.

At the moment, I am working on the development of unsupervised 3D reconstruction from single frames and videos. All my research has been devoted on the development and incorporation of structured layers inside deep neural networks. As a structured layer, we can define an energy based formulation that incorporates human knowledge inside an agnostic machine learning solution. As an undergrad, I worked on the development of human emotion recognition from conversational speech data.

In addition to doing research, I love dancing (tango, salsa) and working out.

For more information about me, see my resumé.

  • Machine Learning
  • Computer Vision
  • Optimization
  • 3D reconstruction
  • Self-supervised Learning
  • Ph.D. in Deep Learning for 3D Reconstruction, 2019

    University College London (UCL)

  • Ph.D. in Deep Learning for Inverse Problems in Computer Vision (transferred to UCL), 2017

    Skoltech University

  • Diploma in Electrical Engineering and Computer Science, 2011

    National Technical University of Athens


Research Intern
Jul 2021 – Sep 2021 Cambridge
Research Intern (part-time)
Oct 2020 – May 2021 London
Research Intern
Jul 2019 – Oct 2019 London
Machine Learning Engineer
Aug 2016 – Aug 2017
Research Assistant
Jan 2016 – Aug 2017
Data Analysis Intern (exchange program)
Jun 2016 – Sep 2016

Recent Publications

Quickly discover relevant content by filtering publications.
(2020). Microscopy Image Restoration with Deep Wiener-Kolmogorov Filters. Computer Vision – ECCV 2020.

PDF Cite Code Project Supplementary

(2019). Iterative Joint Image Demosaicking and Denoising Using a Residual Denoising Network. IEEE Transactions on Image Processing.


(2019). Iterative residual cnns for burst photography applications. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.

PDF Cite

(2019). Pixel Adaptive Filtering Units.

PDF Cite

(2018). Deep Image Demosaicking using a Cascade of Convolutional Residual Denoising Networks. The European Conference on Computer Vision (ECCV).

PDF Cite

(2017). Tweester at SemEval-2017 Task 4: Fusion of Semantic-Affective and pairwise classification models for sentiment analysis in Twitter. Proceedings of the 11th International Workshop on Semantic Evaluation (SemEval-2017).


(2017). Structural Attention Neural Networks for improved sentiment analysis. Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 2, Short Papers.

PDF Cite

(2016). Tweester at SemEval-2016 Task 4: Sentiment Analysis in Twitter Using Semantic-Affective Model Adaptation. Proceedings of the 10th International Workshop on Semantic Evaluation (SemEval-2016).