A computer can look at, and learn from, many more images than a human specialist. AI systems are rapidly becoming indispensable for medical and biological applications. But they still have to learn how to explain their decisions.

With the dramatic growth of the image data on the web, there is an increasing demand of the algorithms capable of understanding the visual information automatically.

The invention of neural networks marks a critical milestone in the pursuit of true artificial intelligence. Despite their impressive performance on various tasks, these networks face limitations in learning efficiently as they are often trained from scratch.

This thesis mainly focuses on multimodal understanding and Visual Question Answering (VQA) via deep learning methods. For technical contributions, this thesis first focuses on improving multimodal fusion schemes via multi-stage vision-language interactions.

This thesis mainly focuses on cross-modal retrieval and single-modal image retrieval via deep learning methods, i.e. by using deep convolutional neural networks.

In this thesis, we investigate how deep generative models (AI models that can create new designs) can be better utilized for technical and industrial design. Our goal is to generate designs that are not only realistic (plausible), but also efficient, reliable, and usable in complex 2D and 3D applica…

In this thesis, deep learning is studied from a statistical perspective. Convergence rates for the worst case risk bounds of neural network estimators are obtained in the classification, density estimation and linear regression model.

In 2018, the number of mobile phone users will reach about 4.9 billion. Assuming an average of 5 photos taken per day using the built-in cameras would result in about 9 trillion photos annually.

This study aims to provide a long time perspective of human landscape manipulation. Studying the roles of prehistoric foragers in past ecosystems is of great importance to establish the character of past 'natural' landscapes and to enhance the management of current ones.

Tomography is a powerful technique to non-destructively determine the interior structure of an object.Usually, a series of projection images (e.g.\ X-ray images) is acquired from a range of different positions.

This project studies the roles of prehistoric foragers in past ecosystems to establish the character of past “natural” landscapes and enhance the management of current ones.

Image segmentation is a fundamental task in computer vision, with applications ranging from medical diagnostics to archaeological research.

Vietnam’s record of expanding access to education, and especially its performance on international assessments such as PISA, has raised questions about what Vietnam got right, how, and why and what insights Vietnam’s experiences might offer for efforts at improving the performance of education systems…

We are living in an information era where the amount of image and video data increases exponentially.

On dinsdag 26 juni 2018, Linlin Sun defended her doctoral thesis ‘International Environmental Obligations and Liabilities in Deep Seabed Mining’. The doctoral research was supervised by Prof. dr. N.J. Schrijver en Prof. dr. E.C.P.D.C. De Brabandere.

Simulating human language understanding on the computer is a great challenge. A way to approach it is to represent natural language meanings in logic, and to use logical provers to determine what does and does not follow from a text. What logic is best to use and how natural language meanings are best…

In this dissertation (titled: Exploring the Synergies between Transfer in Reinforcement Learning and Procedural Content Generation) we explore how the two research fields named in the title, namely Transfer in Reinforcement Learning (TRL) and Procedural Content Generation (PCG) can synergize togethe…

A new microtonal approach of the bass clarinet, to further develope the instrument’s capability to produce not only exact quartertones, but also smaller units.

The recently introduced nature-originated deep eutectic solvents, so-called natural deep eutectic solvents (NADES) are considered as truly green solvents, which composed of natural ingredients found abundantly in organisms.

In deep reinforcement learning, searching and learning techniques are two important components. They can be used independently and in combination to deal with different problems in AI, and have achieved impressive results in game playing and robotics. These results have inspired research into artificial…

With the emergence of online social networks (OSNs), the way people create and share information has changed, which becomes faster and broader than traditional social media.

Building a Deep Cross-media Transfer Network to extract visual relations that relieve the problem of insufficient training data for visual tasks.

Generally the data from remote sensing surveys - the scanning of the earth by satellite or aircraft in order to obtain information about it - is screened manually in archaeology. However, constant monitoring of the earth's surface causes a huge influx of data of high complexity and high quality. To…

The skin has recently gained attention within body studies for its many specific cultural and social associations, in addition to its biology. This project aims to examine the different layers of meaning and the functions invested in the skin in ancient Greece: how did ancient Greek literary and medical…

Implementation of the new sound synthesis and analysis method in music composition.

A brief review essay on some of the work that has been recently published in the emergent field of digital ethnography.

Light reflects from a mirror, but where exactly does this reflection happen? Well, it depends, Martin van Exter and Corné Koks discovered. Their precise calculations, published in Optics Express, are important for designing optical cavities for quantum communication.

This thesis aims to enhance our understanding of galaxies by testing theoretical models of galaxy formation against observations, particularly in the cases of extreme systems which have been found to have an excess of baryonic mass in their central regions, in the form of either supermassive black holes…

In the 1960s at MIT, cognitive scientist Marvin Minsky told a couple of graduate students to program a computer to perform the simple task of recognising objects in pictures, thinking it would be a nice summer project. Scientists from Leiden and the rest of the world are still working on it today.

On Tuesday 26 June Linlin Sun will defend her dissertation entitled ‘International Environmental Obligations and Liabilities in Deep Seabed Mining’. The dissertation has been written under the supervision of prof. dr. N.J. Schrijver and prof. dr. E.C.P.D.C. De Brabandere.

In this thesis we explore machine and deep learning approaches that address keychallenges in high dimensional problem areas and also in improving accuracy in wellknown problems. In high dimensional contexts, we have focused on computational fluid dynamics (CFD) simulations.

Alice Twemlow has recently been interviewed for Design Exhibition Scotland.

Bringing the theoretical world of mathematics and computer science to more applied research areas. That is what Daan Pelt, assistant professor at the Leiden Institute of Advanced Computer Science (LIACS) is trying to achieve with the methods he is developing as a solution to challenges in image proc…

In this thesis, we have studied cell images from two types of cells, including pollen grains and the immune cells, neutrophils. These images are captured using a bright field microscope and a confocal microscope.

Radiography is an important technique to inspect objects, with applications in airports and hospitals. X-ray imaging is also essential in industry, for instance in food safety checks for the presence of foreign objects.

Sewer pipes are an essential infrastructure in modern society and their proper operation is important for public health. To keep sewer pipes operational as much as possible, periodical inspections for defects are performed.

This thesis investigates the application of deep learning techniques in image analysis across various domains, focusing on four main themes: feature extraction, classification, segmentation, and integration, demonstrating the transformative potential of these technologies.

Over several decades, a variety of computational methods for drug discovery have been proposed and applied in practice. With the accumulation of data and the development of machine learning methods, computational drug design methods have gradually shifted to a new paradigm, i.e. deep learning methods…