Computer Vision Lab

Warsaw University of Technology

Computer Vision Lab is a research group at Warsaw University of Technology that gathers faculty and students working together on topics at the intersection of computer vision, machine learning and perception.

Group members

Faculty

Avatar

Kamil Deja

Assistant Professor

Avatar

Tomasz Trzciński

Principal Investigator

PhD Students

Avatar

Michał Bortkiewicz

PhD Student

Avatar

Patryk Będkowski

PhD Student

Avatar

Konrad Cop

PhD Student

Avatar

Bartosz Cywiński

PhD Student

Avatar

Joanna Kaleta

PhD Student

Avatar

Valeriya Khan

PhD Student

Avatar

Daniel Marczak

PhD Student

Avatar

Stanisław Pawlak

PhD Student

Avatar

Paweł Skierś

PhD Student

Avatar

Łukasz Staniszewski

PhD Student

Avatar

Filip Szatkowski

PhD Student

Avatar

Bartosz Świrta

PhD Student

Alumni

Avatar

Jan Dubiński

Alum

Avatar

Kacper Kania

Alum

Avatar

Wojciech Masarczyk

Alum

Avatar

Monika Wysoczańska

Alum

Publications

Towards More Realistic Membership Inference Attacks on Large Diffusion Models

January 2024 WACV 2024
Generative diffusion models, including Stable Diffusion and Midjourney, can generate visually appealing, diverse, and high-resolution …
PDF
Towards More Realistic Membership Inference Attacks on Large Diffusion Models

Bucks for Buckets (B4B): Active Defenses Against Stealing Encoders

December 2023 NeurIPS 2023
Machine Learning as a Service (MLaaS) APIs provide ready-to-use and high-utility encoders that generate vector representations for …
PDF
Bucks for Buckets (B4B): Active Defenses Against Stealing Encoders

BlendFields: Few-Shot Example-Driven Facial Modeling

June 2023 CVPR 2023
Generating faithful visualizations of human faces requires capturing both coarse and fine-level details of the face geometry and …
PDF Project Video Project Page
BlendFields: Few-Shot Example-Driven Facial Modeling

Learning Data Representations with Joint Diffusion Models

April 2023 ECML 2023
Joint machine learning models that allow synthesizing and classifying data often offer uneven performance between those tasks or are …
PDF Code
Learning Data Representations with Joint Diffusion Models

On Analyzing Generative and Denoising Capabilities of Diffusion-based Deep Generative Models

December 2022 NeurIPS 2022
Diffusion-based Deep Generative Models (DDGMs) offer state-of-the-art performance in generative modeling. Their main strength comes …
PDF Code
On Analyzing Generative and Denoising Capabilities of Diffusion-based Deep Generative Models
See all publications

Funded Projects

Continual Self-Supervised Representation Learning - Grant PRELUDIUM BIS funded by National Science Centre Poland

We tackle the problem of learning representations in continual learning regime.
Continual Self-Supervised Representation Learning - Grant PRELUDIUM BIS funded by National Science Centre Poland

OPUS funding from National Science Centre in Poland

Application of deep generative models for lifelong learning.
OPUS funding from National Science Centre in Poland

Microsoft Research PhD Scholarship Award

Low shot realistic human rendering from partial information.
Microsoft Research PhD Scholarship Award

Grant of Priority Research Domain at WUT - Artificial Intelligence and Robotics

Binary representations and their application in continual learning.
Grant of Priority Research Domain at WUT - Artificial Intelligence and Robotics

Grant of Priority Research Domain at WUT - High Energy Physics and Experimental Techniques

WUT@ALICE: Study of fundamental properties of strongly interacting matter with particle correlations and machine learning in ALICE at LHC.
Grant of Priority Research Domain at WUT - High Energy Physics and Experimental Techniques
See all

Contact