Research - NAVER LABS Europe
preloder

RESEARCH

Our technical expertise is enriched with a deep understanding of how people interact with technology.

Some of the big cross-disciplinary challenges we address:

Representation learning: turning all data entities (text, images, videos, web pages, emails, …) into semantically-rich representations to that can be efficiently processed, mined and related to each other.

Interactive learning: including the human in the learning loop to benefit from the complementary capabilities of humans and machines

Lifelong learning: exploring how to learn and manage prediction systems, policies or processes in a sustainable manner, i.e. in a manner that makes maximum reuse of previous experiences.

Machine reasoning: connecting existing pieces of knowledge to create new knowledge, e.g. being able to answer complex questions that require multiple supporting facts.

Decision making: learning how to make actionable decisions in uncertainty, especially in ever-changing environments

Agility: we live in an ever-changing world which means the technology we use must evolve in synergy with the people and business processes that touch it.

User centric: an important challenge in any technology is to the adoption by people. Only then does science become innovation. Our solutions are grounded in a deep understanding of how people work and how technology can help them every day.

Collaboration

Our scientists collaborate with national and international partners, academics and businesses to solve problems and invent technology and services that will have impact in the real world.

Discover how our research is recognized.

Research illustrating image woman and 2 man in front of computers
, , , , , , , , ,
12 June 2019

“I Don’t Understand…” Issues in Self-Quantifying Commuting

This paper illustrates the results of a study that we conducted on the effectiveness of self-tracking of commuting data where participants received daily feedback on the financial costs and CO2 emissions associated to their mobility practices.
29 May 2019

Improving the Generalization of Visual Navigation Policies using Invariance Regularization

Training agents to operate in one environment often yields overfitted models that are unable to generalize to the changes in that environment.
27 May 2019

Visual Localization by Learning Objects-of-Interest Dense Match Regression

We introduce a novel CNN-based approach for visual localization from a single RGB image that relies on densely matching a set of Objects-of-Interest (OOIs).
12 June 2019

NAVER LABS Europe ranks first in international benchmark in visual localization.

Results to be presented at Long Beach, California on June 17th at CVPR 2019, the world’s premier event in computer vision.
28 May 2019

Thirty-sixth International Conference on Machine Learning (ICML 2019), Long Beach, CA, USA

NAVER LABS EUROPE and Clova AI, are Gold Sponsors of the event. We're presenting on the booth in between workshops and papers and hope to see you there!
28 May 2019

CVPR 2019, Long Beach, CA, USA

We have 3 papers, 2 workshop papers and are winners of the 'Local Features' Visual Localization challenge! Come talk to us on the booth.
,
14 June 2019
Augmented Reality Guide cover image

An Augmented Reality Guide for Museums

The acme of perfection of AR is when you can’t distinguish the virtual from the real. We experiment with AI and computer vision in our AR museum guide ARAO to make the physical experience a natural one.
14 June 2019
Making maps evergreen with deep learning, robots and computer vision

Making maps evergreen with deep learning, robots and computer vision

A system that correctly detects when places have changed to automatically update complex indoor maps. Dataset available for research.
7 June 2019
MARS_CVPR blog image

MARS: Motion-Augmented RGB Stream for Action Recognition

This blog presents our CVPR’19 paper on “MARS: Motion-Augmented RGB Stream for Action Recognition” done with the Thoth team at Inria.

Related content