Lifeboat Foundation

Wenn wir unsere Augen öffnen, dann fällt es uns ganz leicht, die verschiedene Objekte, Menschen und Tiere um uns herum zu sehen. Bisher war die weitreichende Forschungsmeinung, dass ein ganz wesentliches Ziel unserer Wahrnehmung ist, Objekte zu erkennen und verschiedenen Kategorien zuzuordnen – zum Beispiel, ob dieses Objekt vor uns ein Hund ist und ob ein Hund zur Kategorie der Tiere zählt. Forschende vom Max-Planck-Institut für Kognitions-und Neurowissenschaften in Leipzig und der Justus-Liebig-Universität Gießen in Zusammenarbeit mit den National Institutes of Health in den USA konnte nun zeigen, dass dieses Bild unvollständig ist. In einer aktuellen Studie im Fachjournal Nature Human Behaviour schreiben sie, dass sich die Hirnaktivität beim Sehen von Objekten viel besser mit einer Vielzahl verhaltensrelevanter Dimensionen erklären lässt.

Bisher dachte man, dass das visuelle System in unserem Gehirn die gesehenen Objekte in sehr grundlegende Merkmale zerlegt und dann nach und nach wieder zusammensetzt, mit dem Ziel, deren Erkennen zu ermöglichen. „Unsere Ergebnisse haben gezeigt, dass Erkennen und Kategorisieren zwar wichtige Ziele unseres Sehens sind, aber bei weitem nicht die einzigen.“, sagt Letztautor Martin Hebart, Gruppenleiter am MPI CBS und Professor an der Justus-Liebig-Universität. „Tatsächlich finden wir verhaltensrelevante Signale an allen Verarbeitungsstufen im visuellen System. Dies konnten wir aus der Analyse der von uns entdeckten verhaltensrelevanten Dimensionen ableiten.” Im Vorfeld hatten die Forscher mit einem Computermodell aus Verhaltensdaten von über 12.000 Studienteilnehmer*innen 66 Objektdimensionen identifiziert.

Given the right conditions, certain types of cells are able to self-assemble into new lifeforms after the organism they were once part of has died.

In polymers, the competition between the folding and aggregation of chains, both at an individual level and between chains, can determine the mechanical, thermal, and conductive properties of such materials. Understanding the interplay of folding and aggregation presents a significant opportunity for the development and discovery of polymeric materials with tailored properties and functionalities.

Google researchers have created an innovative AI model called Health Acoustic Representations (HeAR), designed to identify acoustic biomarkers for diseases like tuberculosis.

It can listen to human sounds and flag early signs of disease.

Continue reading “New AI model helps researchers detect disease based on coughs” »

Are we all just part of an endless cosmic cycle?

Hydrogen has been defined on numerous occasions as “the fuel of the future”. We have seen other alternatives, such as ammonia or even methanol (which you may remember meeting with us), but what if there was an even more powerful one? Hawking predicted decades ago that the most powerful one could exist, and now they have finally created it. This is the new engine that has everything to revolutionize the planet but would require a huge mobilization of resources to manufacture.

The idea of using thorium for fueling cars has created the immense interest from auto enthusiasts, as such cars may become a clean, efficient and almost inexhaustible energy source for transport in the future. Nevertheless, the prospects of this technology are not as simple as may be suggested by this example, and at the moment, this technology is still rather hypothetical.

A thorium-powered car engine concept is based on the use of the radioactive material known as thorium as fuel. In principle, this engine employed a tiny measure of thorium to release heat through nuclear fission, and the heat was further transformed into electricity to run the car.

As its name implies, the Bose glass exhibits certain glass-like properties, with all particles in the system becoming localized. This means that each particle remains confined to its position, without interacting or blending with its neighbors.

If coffee behaved in this way, for example, stirring milk into it would result in a permanent pattern of black and white stripes that never mix into a uniform color.

In a localized system like the Bose glass, particles don’t mix with their environment, which suggests that quantum information stored within such a system could be retained for much longer periods. This property has significant implications for quantum computing and information storage.

The team was able to recreate NASA’s iconic “Pillars of Creation” image from the James Webb Space Telescope data — originally created by a supercomputer — using just a tabletop computer.

An Open-Source Library for Gaussian Splatting discuss:

Gsplat is an open-source library designed for training and developing Gaussian Splatting methods.

Join the discussion on this paper page.