Thomas TANAY

Senior Research scientist at Huawei Technologies Ltd, Noah's Ark Lab (London, UK).

Selected publications

GRVS: A Generalizable and Recurrent Approach to Monocular Dynamic View Synthesis.
Tanay*, Brahimi*, Nazarczuk, Zhang, Catley‑Chandar, Moreau, Zhang and Pérez‑Pellitero
arXiv, 2025
Project page arXiv
Charge: A Comprehensive Benchmark and Dataset for Dynamic Novel View Synthesis.
Nazarczuk, Tanay, Zhang and Pérez‑Pellitero
arXiv, 2025
Project page Dataset
Better Together: Unified Motion Capture and 3D Avatar Reconstruction.
Moreau, Brahimi, Shaw, Papaioannou, Tanay, Zhang and Pérez‑Pellitero
arXiv, 2025
arXiv
ViDAR: Video Diffusion‑Aware 4D Reconstruction from Monocular Inputs.
Nazarczuk*, Catley‑Chandar*, Tanay, Zhang, Slabaugh and Pérez‑Pellitero
NeurIPS, 2025
Project page arXiv Code
Global Latent Neural Rendering.
Tanay and Maggioni
CVPR, 2024
Project page arXiv
Adaptive Spiral Layers for Efficient 3D Representation Learning on Meshes.
Babiloni, Maggioni, Tanay, Deng, Leonardis and Zafeiriou
ICCV, 2023
CVF Code
Efficient View Synthesis and 3D-based Multi-Frame Denoising with Multiplane Feature Representations.
Tanay, Leonardis, and Maggioni
CVPR, 2023
Project page arXiv
Tunable Convolutions with Parametric Multi-Loss Optimization.
Maggioni, Tanay, Babiloni, McDonagh, and Leonardis
CVPR, 2023
arXiv
Diagnosing and Preventing Instabilities in Recurrent Video Processing.
Tanay, Sootla, Maggioni, Dokania, Torr, Leonardis, and Slabaugh
TPAMI, 2022
arXiv TPAMI
FlexHDR: Modelling Alignment and Exposure Uncertainties for Flexible HDR Imaging.
Catley‑Chandar, Tanay, Vandroux, Leonardis, Slabaugh, and Pérez‑Pellitero
TIP, 2022
Project page arXiv TIP
Batch Normalization is a Cause of Adversarial Vulnerability.
Galloway, Golubeva, Tanay, Moussa and Taylor
arXiv, 2019
arXiv
Adversarial Training versus Weight Decay.
Galloway, Tanay and Taylor
arXiv, 2018
arXiv
A Boundary Tilting Persepective on the Phenomenon of Adversarial Examples.
Tanay and Griffin
arXiv, 2016
arXiv
HeX and the Single Anthill: Playing Games with Aunt Hillary.
Bishop, Nasuto, Tanay, Roesch and Spencer
Fundamental Issues of Artificial Intelligence, 2016

Workshops and Challenges

Sparse Neural Rendering Challenge. (Organizer)
Nazarczuk, Tanay, Catley-Chandar, Shaw, Timofte, PĂ©rez-Pellitero
Advances in Image Manipulation workshop, ECCV, 2024
Webpage Dataset and Benchmark Methods and Results
MPFER‑H: Multiplane Features Encoder‑Renderer for Human Heads. (Ranked #1)
Tanay and Maggioni
To NeRF or not to NeRF: a View Synthesis Challenge for Human Heads, ICCV, 2023
Webpage CVF CodaLab
Factorized Dynamic Fully Connected Layers for Neural Networks.
Babiloni, Tanay, Deng, Maggioni, and Zafeiriou.
Workshop on Resource Efficient Deep Learning for Computer Vision, ICCV, 2023
CVF
Multi‑scale network with deformable temporal fusion for compressed video restoration.
Song, Huang, Maggioni, Fu, Xiao, Li and Tanay
NTIRE 2021 Challenge on Quality Enhancement of Compressed Video, CVPR, 2021
CVF
Learnable Non‑local Image Denoising.
Maggioni, Marras, Tanay, Slabaugh and Yan
NTIRE 2020 Challenge on Real Image Denoising, CVPR, 2020
CVF
Built‑in Vulnerabilities to Imperceptible Adversarial Perturbations. (Oral)
Tanay, Andrews and Griffin
BMVA Technical Meeting, 2019
arXiv

Posts

How Government Expenditures Finance Themselves.
A macroeconomic model from Wynne Godley and Marc Lavoie.
2022
A New Angle on L2 Regularization.
An explorable explanation of adversarial examples in linear classification.
2018
KDnuggets
Face Verification and Clustering.
2015
Stochastic Diffusion Search Applied to Trees.
2013