Global ETD Search
Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.
Results
Showing 1 to 20 of 135 for “"world model"”.
-
World Model Online Evaluation
… and informative data for training perception models. Active learning is based on the principle that exposing models to previously unseen or challenging data improves their performance. To facilitate this, this thesis evaluates several motion-based methods for detecting inconsistencies in the …
-
Graph-based world-model for robotic manipulation
… that supports this functionality is a robust world model to act as a repository of environmental information. The research community has provided many world-modeling solutions to support autonomous vehicle navigation. As such, they focus primarily on preventing collisions with the environment. …
-
Dynamic world model for context-aware environments
… of this dissertation is a dynamic symbolic world model which is connected to sensor networks through data acquisition and processing modules. Being dynamic, it is capable of accompanying the state of the represented spaces in real time. This model can be distributed over several servers as …
-
World Model Uncertainty for Sample Efficient Model-Based Reinforcement Learning
… of Reinforcement Learning (RL) agents on real-world robots, Model-Based Reinforcement Learning (MBRL) addresses data inefficiency by using a world model to simulate the real environment, thereby minimising wear and tear on mechanical components. Uncertainty has emerged as a metric for …
-
World model based multi-agent proximal policy optimization framework for multi-agent pathfinding
… by the laws of physics. Within the framework, a World model is utilized to extract and abstract representation features from the global map, leveraging the global context to enhance the training process. This approach involves decoupling the feature extractor from the agent training process, …
-
Toward Generalist Ophthalmic Intelligence: From Scalable Multimodal Representation Learning to a Clinical World Model
… dimensionalities, and time points, and models built for one configuration degrade on the next. Prior work has largely responded by narrowing scope to one modality, one task and one time point, which makes each problem tractable and their union intractable. This dissertation argues that …
-
A World Model Reinforcement Learning Approach for Vehicle Control: A vehicle controller based on DreamerV3 and a research platform for training and evaluating machine learning based controllers
… of this thesis was to investigate whether a world-model-based reinforcement learning approach could be used for path following on a small off-road unmanned ground vehicle. The method involved building a 1/10 scale RC car platform equipped with a binocular camera and developing a ROS 2 graph …
-
Tracing Buddhist Responses to the Crisis of Cosmography
… and regions, traditionally posited a similar world model — one that is flat and centered by giant Mount Meru. This world model is chiefly featured in Vasubandhu’s fourth century CE text, the Abhidharmakośabhāṣyam. In 1552, Christian missionary Francis Xavier introduced European …
-
A Deep Hierarchical Variational Autoencoder for World Models in Complex Reinforcement Learning Environments
Model-based reinforcement learning (MBRL) approaches leverage learned models of the environment to plan and make optimal decisions, reducing the need for extensive real-world interactions and enabling more efficient learning in complex domains such as robotics, autonomous systems, and resource …
-
Learning to Reach Goals from Suboptimal Demonstrations via World Models
… to learn faster. In this thesis, we propose World Model Contrastive Reinforcement Learning (WM-CRL), which augments CRL with representations from a world model (WM). The world model is trained to anticipate future state embeddings from past state–action pairs, thereby encoding the dynamics of …
-
A STOCHASTIC POSITION SENSING SYSTEM BASED ON MACHINE VISION AND WHEEL ENCODERS
… robot. The fused sensor data are used to build a World Model or map; since the fused sensor data are stochastic then the features (parameters, objects) in the World Model are also stochastic. The position is then determined by the relative position of features with respect to the World Model. The …
-
Kosminio žingsninio strateginio žaidimo kūrimas. Game Engine projektavimas ir realizavimas /
The purpose of work is to create part of world model module which is associated with resource creation. In the course of this work requirements for this part of module were analyzed and an implementation project was made. Using this project a part of world model was created, which is responsible …
-
Statistical characterization of sensors for sensor fusion
… advanced telerobotic systems to use a detailed model of the operating environment to plan tasks and drive simulations. How well these systems operate is highly dependent upon the accuracy of this model. As objects are moved by manipulators, inaccuracies in the world model will accumulate, …
-
LEARNING TO ACT FROM DIVERSE DATA SOURCES VIA WORLD MODELS
… learn generalist agents? I argue that learning world models that predict future outcomes of actions directly from image observations is a uniquely suitable approach for training generalizable agents. I will discuss how world models can enable powerful unsupervised exploration and how to use a …
-
Continuum: an architecture for user evolvable collaborative virtual environments
… Continuum's architecture supplies a world model and defines how to combine object state, behavior code, and resource data into this single shared structure. The system frees distributed users from the constraints of monolithic centralized virtual world architectures and instead allows …
Page 1 of 7