Lucas Theis

I am a senior research scientist at Google DeepMind working on compression using neural networks. My general interests are around the application of probabilistic machine learning, information theory, and statistics to challenging real-world problems.

Before joining Google I worked at Twitter where I developed models for compression, super-resolution, topic modeling, image cropping, and other tasks. I joined Twitter through the acquisition of Magic Pony Technology, a London based startup working on compression. Before that I did my PhD at the Max Planck Research School for Neural Information Processing in Tübingen, working in the lab of Matthias Bethge on natural image statistics and computational neuroscience.

Resume lucas@theis.io

On the web

CrossValidated GitHub Twitter Facebook Flickr Pinboard

Publications

	L. Theis What makes an image realistic? arXiv:2403.04493, 2024 #perceptual quality #realism #compression #generative modeling #outlier detection URL RIS BibTex
	H. Kim, M. Bauer, L. Theis, J. R. Schwarz, and E. Dupont C3: High-performance and low-complexity neural compression from a single image or video Computer Vision and Pattern Recognition, 2024 #compression Code URL Project RIS BibTex
	D. Severo, L. Theis, and J. Ballé The Unreasonable Effectiveness of Linear Prediction as a Perceptual Metric International Conference on Learning Representations, 2024 #perceptual quality #bapps #lpips #lasi #ssim URL RIS BibTex
	E. Hoogeboom, E. Agustsson, F. Mentzer, L. Versari, G. Toderici, and L. Theis High-Fidelity Image Compression with Score-based Generative Models arXiv:2305.18231, 2023 #compression #diffusion #rectified flow URL Files RIS BibTex
	G. Flamich and L. Theis Adaptive Greedy Rejection Sampling IEEE International Symposium on Information Theory, 2023 #channel simulation #compression #information theory URL PDF RIS BibTex
	Y. Yang, S. Mandt, and L. Theis An Introduction to Neural Data Compression Foundations and Trends in Computer Graphics and Vision, 15(2), 113-200, 2023 #compression URL PDF RIS BibTex
	L. Theis, T. Salimans, M. D. Hoffman, and F. Mentzer Lossy Compression with Gaussian Diffusion arXiv:2206.08889, 2022 #compression #diffusion #channel simulation URL RIS BibTex
	A. Shah, W.-N. Chen, J. Balle, P. Kairouz, and L. Theis Optimal Compression of Locally Differentially Private Mechanisms Artificial Intelligence and Statistics, 2022 #differential privacy #compression #channel simulation Code URL RIS BibTex
	L. Theis and N. Yosri Algorithms for the Communication of Samples Proceedings of the 39th International Conference on Machine Learning, 2022 #compression #information theory #channel simulation Code URL PDF RIS BibTex
	L. Theis and J. Ho Importance weighted compression Neural Compression Workshop at ICLR, 2021 #compression #deep learning #bits back URL RIS BibTex
	L. Theis and A. B. Wagner A coding theorem for the rate-distortion-perception function Neural Compression Workshop at ICLR, 2021 #compression #information theory #perceptual quality URL RIS BibTex
	L. Theis and E. Agustsson On the advantages of stochastic encoders Neural Compression Workshop at ICLR, 2021 #compression #information theory URL RIS BibTex
	E. Agustsson and L. Theis Universally Quantized Neural Compression Advances in Neural Information Processing Systems 33, 2020 #compression #deep learning #channel simulation URL PDF Appendix RIS BibTex
	I. Korshunova, H. Xiong, M. Fedoryszak, and L. Theis Discriminative Topic Modeling with Logistic LDA Advances in Neural Information Processing Systems 33, 2019 #lda #bayesian inference #topic modeling Code URL PDF Appendix RIS BibTex
	T. Nguyen-Phuoc, C. Li, L. Theis, C. Richardt, and Y.-L. Yang HoloGAN: Unsupervised learning of 3D representations from natural images International Conference on Computer Vision, 2019 #generative modeling #3d #deep learning Code URL PDF Video RIS BibTex
	K. Storrs, S. V. Leuven, S. Kojder, L. Theis, and F. Huszár Adaptive Paired-Comparison Method for Subjective Video Quality Assessment on Mobile Devices Picture Coding Symposium, 2018 #compression #perceptual quality #psychophysics URL Blog RIS BibTex
	L. Theis, I. Korshunova, A. Tejani, and F. Huszár Faster gaze prediction with dense networks and Fisher pruning arXiv:1801.05787, 2018 #pruning #fisher information #saliency URL Blog RIS BibTex
	I. Korshunova, W. Shi, J. Dambre, and L. Theis Fast Face-swap Using Convolutional Neural Networks International Conference on Computer Vision, 2017 #face-swap #cagenet #deep learning URL RIS BibTex
	C. Ledig, L. Theis, F. Huszar, J. Caballero, A. Aitken, A. Tejani, J. Totz, Z. Wang, et al. Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network Computer Vision and Pattern Recognition, 2017 #super-resolution #deep learning URL RIS BibTex
	L. Theis, W. Shi, A. Cunningham, and F. Huszár Lossy Image Compression with Compressive Autoencoders International Conference on Learning Representations, 2017 #compression #deep learning URL PDF Poster Files RIS BibTex
	C. Sønderby, J. Caballero, L. Theis, W. Shi, and F. Huszár Amortised MAP Inference for Image Super-resolution International Conference on Learning Representations, 2017 #super-resolution #deep learning URL RIS BibTex
	L. Theis, P. Berens, E. Froudarakis, J. Reimer, M. Roman-Roson, T. Baden, T. Euler, A. S. Tolias, et al. Benchmarking spike rate inference in population calcium imaging Neuron, 90(3), 471-482, 2016 #two-photon imaging #spiking neurons Code URL DOI PDF Poster RIS BibTex
	L. Theis, A. van den Oord, and M. Bethge A note on the evaluation of generative models International Conference on Learning Representations, 2016 #generative modeling Code URL PDF Talk RIS BibTex
	L. Theis and M. Bethge Generative Image Modeling Using Spatial LSTMs Advances in Neural Information Processing Systems 28, 2015 #deep learning #generative modeling #natural image statistics #lstm #mcgsm Code URL PDF Supplemental Poster RIS BibTex
	L. Theis and M. D. Hoffman A trust-region method for stochastic variational inference with applications to streaming data Proceedings of the 32nd International Conference on Machine Learning, 2015 #lda #streaming #svi #bayesian inference #topic modeling Code URL PDF Supplemental Poster Talk RIS BibTex
	M. Kümmerer, L. Theis, and M. Bethge Deep Gaze I: Boosting Saliency Prediction with Feature Maps Trained on ImageNet ICLR Workshop, 2015 #saliency #deep learning URL PDF RIS BibTex
	H. E. Gerhard, L. Theis, and M. Bethge Modeling Natural Image Statistics Biologically-inspired Computer Vision—Fundamentals and Applications, Wiley VCH, 2015 #natural image statistics #mcgsm #ica #psychophysics URL ISBN PDF RIS BibTex
	S. Sra, R. Hosseini, L. Theis, and M. Bethge Data modeling with the elliptical gamma distribution Artificial Intelligence and Statistics, 2015 #density estimation #natural image statistics #mixture modeling URL PDF RIS BibTex
	A. M. Chagas, L. Theis, B. Sengupta, M. Stüttgen, M. Bethge, and C. Schwarz Functional analysis of ultra high information rates conveyed by rat vibrissal primary afferents Frontiers in Neural Circuits, 7(190), 2013 #neuroscience #spiking neurons URL RIS BibTex
	L. Theis, A. M. Chagas, D. Arnstein, C. Schwarz, and M. Bethge Beyond GLMs: A Generative Mixture Modeling Approach to Neural System Identification PLoS Computational Biology, 9(11), 2013 #generalized linear model #spiking neurons #mixture models #generative modeling Code URL DOI PDF RIS BibTex
	L. Theis, J. Sohl-Dickstein, and M. Bethge Training sparse natural image models with a fast Gibbs sampler of an extended state space Advances in Neural Information Processing Systems 25, 2012 #natural image statistics #ica #overcompleteness #bayesian inference Code PDF Supplemental Poster RIS BibTex
	L. Theis, R. Hosseini, and M. Bethge Mixtures of Conditional Gaussian Scale Mixtures Applied to Multiscale Image Representations PLoS ONE, 7(7), 2012 #natural image statistics #random fields #mcgsm #mixture models Code URL DOI PDF RIS BibTex
	L. Theis, S. Gerwinn, F. Sinz, and M. Bethge In All Likelihood, Deep Belief Is Not Enough Journal of Machine Learning Research, 12, 2011 #natural image statistics #deep belief networks #boltzmann machines #deep learning Code PDF RIS BibTex