Luke Melas-Kyriazi

Luke Melas-Kyriazi


I'm a Rhodes Scholar and PhD student at Oxford University. My research spans machine learning, computer vision, and natural language processing. I am lucky to be advised by Professor Andrea Vedaldi in the Visual Geometry Group (VGG). I'm always interested in meeting new people and hearing about potential collaborations. If you'd like to get in touch with me, please email me at lukemk [at] robots [dot] ox [dot] ac [dot] uk.


~ News ~


~ Publications ~

IM-3D: Iterative Multiview Diffusion and Reconstruction for High-Quality 3D Generation

L Melas-Kyriazi, I Laina, C Rupprecht, N Neverova, A Vedaldi, O Gafni, F Kokkinos

ICML 2024

Fixed Point Diffusion Models

X Bai*, L Melas-Kyriazi*

CVPR 2024

GES: Generalized Exponential Splatting for Efficient Radiance Field Rendering

A Hamdi, L Melas-Kyriazi, G Qian, J Mai, R Liu, C Vondrick, B Ghanem, A Vedaldi

CVPR 2024

A Benchmark for Learning to Translate a New Language from One Grammar Book

G Tanzer, M Suzgun, E Visser, D Jurafsky, L Melas-Kyriazi

ICLR 2024 (Spotlight)

RealFusion: 360-Degree Reconstruction of Any Object from a Single Image

L Melas-Kyriazi, C Rupprecht, I Laina, A Vedaldi

CVPR 2023

PC^2: Projection-Conditioned Point Cloud Diffusion for Single-Image 3D Reconstruction

L Melas-Kyriazi, C Rupprecht, A Vedaldi

CVPR 2023 (Highlight)

The Harvard USPTO Patent Dataset: A Large-Scale, Well-Structured, and Multi-Purpose Corpus of Patent Applications

M Suzgun, L Melas-Kyriazi, S Sarkar, SD Kominers, S Shieber

NeurIPS 2023 - Datasets and Benchmarks (Spotlight)

Understanding Self-Supervised Features for Learning Unsupervised Instance Segmentation

P Engstler*, L Melas-Kyriazi*, C Rupprecht, I Laina

NeurIPS 2023 - Workshop on Self-Supervised Learning

Augmenting Medical Image Classifiers with Synthetic Data from Latent Diffusion Models

LW Sagers*, JA Diao*, L Melas-Kyriazi*, M Groh, P Rajpurkar, AS Adamson, V Rotemberg, R Daneshjou, AK Manrai

In Submission

CVMedical ML

Follow the Wisdom of the Crowd: Effective Text Generation via Minimum Bayes Risk Decoding

M Suzgun, L Melas-Kyriazi, D Jurafsky

ACL Findings 2023

Intrinsic Gradient Compression for Scalable and Efficient Federated Learning

L Melas-Kyriazi*, F Wang*

ACL 2023 - Workshop on Federated Learning for NLP

Prompt-and-Rerank: A Method for Zero-Shot and Few-Shot Arbitrary Textual Style Transfer with Small Language Models

M Suzgun, L Melas-Kyriazi, D Jurafsky

EMNLP 2022 (Oral)

Deep Spectral Methods: A Surprisingly Strong Baseline for Unsupervised Semantic Segmentation and Localization

L Melas-Kyriazi, C Rupprecht, I Laina, A Vedaldi

CVPR 2022 (Oral)

Finding an Unsupervised Image Segmenter in Each of Your Deep Generative Models

L Melas-Kyriazi, C Rupprecht, I Laina, A Vedaldi

ICLR 2022

Do You Even Need Attention? A Stack of Feed-Forward Layers Does Surprisingly Well on ImageNet

L Melas-Kyriazi

arXiv

ML for Medicine

PixMatch: Unsupervised Domain Adaptation via Pixelwise Consistency Training

L Melas-Kyriazi, AK Manrai

CVPR 2021

Show, Edit and Tell: A Framework for Editing Image Captions

F Sammani*, L Melas-Kyriazi*

CVPR 2020

Prediction of Chronological and Biological Age from Laboratory Data

L Sagers, L Melas-Kyriazi, CJ Patel, AK Manrai

Journal on Aging, 2020

ML for Medicine

Generation-Distillation for Efficient Natural Language Understanding in Low-Data Settings

L Melas-Kyriazi, G Han, C Liang

EMNLP 2019 - Workshop on Deep Learning for Low-Resource NLP

Encoder-Agnostic Adaptation for Conditional Language Generation

ZM Ziegler, L Melas-Kyriazi, S Gehrmann, AM Rush

arXiv

MLNLP

Training for Diversity in Image Paragraph Captioning

L Melas-Kyriazi, G Han, AM Rush

EMNLP 2018