Training diffusion models with federated learning: A communication-efficient model for cross-silo federated image generation

The training of diffusion-based models for image generation is predominantly controlled by a select few Big Tech companies, raising concerns about privacy, copyright, and data authority due to the lack of transparency regarding training data. Hence, we propose a federated diffusion model scheme that enables the independent and collaborative...

Efficient Temporal Action Localization via Vision-Language Modelling: An Empirical Study on the STALE Model's Efficiency and Generalizability in Resource-constrained Environments

Temporal Action Localization (TAL) aims to localize the start and end times of actions in untrimmed videos and classify the corresponding action types. TAL plays an important role in understanding video. Existing TAL approaches heavily rely on deep learning and require large-scale data and expensive training processes. Recent advances in...

Extending CTAB-GAN+ with StackGAN architecture

As privacy regulations (e.g. European General Data Protection Regulation) often prevent valuable flows of data between stakeholders, data synthesis can play a crucial role in sharing captured value in data sets without sharing personal details. Different attempts have been made at solving this problem with Generative Adversarial Networks (GAN)...

Towards Trust in Human-AI Teams

Human-AI teams require trust to operate efficiently and solve certain tasks like search & rescue. Trustworthiness is measured using the ABI model; Ability, Benevolence and Integrity. This research paper tries to observe the effect a conflicting robot has on the human trustworthiness. The hypothesis we try to test is: “human trustworthiness...

Using Human Workload to Adjust Agent Explanations in Human-agent Teamwork

Artificial intelligence systems assist humans in more and more cases. However, such systems' lack of explainability could lead to a bad performance of the teamwork, as humans might not cooperate with or trust systems with black-box algorithms opaque to them. This research attempts to improve the explainability of artificial intelligence systems...

AI Soccer: the Most Effective Methods to Dispossess the Opponent Player

In this paper, one of the challenges that comes with defending in AI soccer is highlighted and an attempt is made in finding a solution for the problem. In soccer, defense is an important part of the game and the research question in this research is formulated as follows: what are the most effective methods to take by surprise and dispossess...

Procedural automation of general game level generation: the good, the bad and the ugly

The monumental goal of Artificial Intelligence (AI) is to model general solutions that can be applied to perform a variety of tasks that normally demand human intelligence to solve. Traditionally human game developers painstakingly design and tweak levels until achieving the precise output of their heart’s desire. In the gaming industry, AI for...

Multi-agent systems in StarCraft

The Interactive Intelligence section of the TU Delft would like to replace the game UT3 that is used in the Multi-Agent Systems project (MAS-project), which is a part of the bachelor curriculum of computer science. The game that has been selected for this purpose is StarCraft. StarCraft is called the `next challenge in AI development' by Google....