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” »
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.
Game Developer jourverse, who is currently working on a tutorial series focused on building a traffic system in Unreal Engine 5, shared a demo project file for this procedural road network integrated with vehicle AI for obstacle avoidance, using A* for pathfinding.
The developer explained that both the A* algorithm and the road editor mode are implemented in C++, with no use of neural networks. Vehicle AI operations like spline following, reversing, and performing 3-point turns are handled through Blueprints. The vehicle AI navigates using two paths: the green spline for the main route and the blue spline for obstacle avoidance. The main spline leverages road network nodes to determine the path to the target via A* on FPathNode, which includes adjacent road nodes.
For obstacle detection, the vehicle employs polynomial regression to predict its future position. Upon detecting an obstacle, a grid of sphere traces is generated to map the obstacle’s location, and another A* algorithm is employed to create a path around the obstacle.
With maps of the connections between neurons and artificial intelligence methods, researchers can now do what they never thought possible: predict the activity of individual neurons without making a single measurement in a living brain.
For decades, neuroscientists have spent countless hours in the lab painstakingly measuring the activity of neurons in living animals to tease out how the brain enables behavior. These experiments have yielded groundbreaking insights into how the brain works, but they have only scratched the surface, leaving much of the brain unexplored.
Now, researchers are using artificial intelligence and the connectome—a map of neurons and their connections created from brain tissue —to predict the role of neurons in the living brain. Their paper has been published in the journal Nature.
Read about the importance of International Observe the Moon Night and how you can celebrate it on September 14, 2024!
Beginning in 2010, NASA began International Observe the Moon Night based on two events occurring simultaneously in 2009 during the International Year of Astronomy celebration: “We’re at the Moon!”, which was sponsored by the Lunar Reconnaissance Orbiter (LRO) and the Lunar Crater Observation and Sensing Satellite (LCROSS) teams, and “National Observe the Moon Night”, which was hosted in the United States.
This year’s International Observe the Moon Night is occurring on September 14 with the goal of sharing the incredible science and wonder of the Moon, including its observational and scientific history, why it’s so important to study, and how we’re studying it. For example, evidence has suggested that ancient humans as far back as 20,000 years ago used the Moon as a timekeeping device due to the changing phases of the Moon over the course of a month. Additionally, when observing the Moon with either the naked eye or a telescope, the Moon’s surface exhibits both bright and dark colors, which are the Moon’s lava plains and highlands, respectively.
Enjoy the videos and music you love, upload original content, and share it all with friends, family, and the world on YouTube.
By tapping into a decades-old mathematical principle, researchers are hoping that Kolmogorov-Arnold networks will facilitate scientific discovery.
An up-and-coming startup in the world of AI chips might be giving Nvidia a run for its money.
NASA’s Valkyrie robot is an intimidating figure. It is currently being put through its paces at the Karda laboratory in Australia so researchers can work out what it would take to get a humanoid robot onto offshore energy facilities or into space. New Scientist‘s James Woodford took the controls to see what the $2 million-plus device is capable of.
Abstract: Recent advancements in large language models (LLMs) have sparked optimism about their potential to accelerate scientific discovery, with a growing number of works proposing research agents that autonomously generate and validate new ideas. Despite this, no evaluations have shown that LLM systems can take the very first step of producing novel, expert-level ideas, let alone perform the entire research process. We address this by establishing an experimental design that evaluates research idea generation while controlling for confounders and performs the first head-to-head comparison between expert NLP researchers and an LLM ideation agent. By recruiting over 100 NLP researchers to write novel ideas and blind reviews of both LLM and human ideas, we obtain the first statistically significant conclusion on current LLM capabilities for research ideation: we find LLM-generated ideas are judged as more novel (p < 0.05) than human expert ideas while being judged slightly weaker on feasibility. Studying our agent baselines closely, we identify open problems in building and evaluating research agents, including failures of LLM self-evaluation and their lack of diversity in generation. Finally, we acknowledge that human judgements of novelty can be difficult, even by experts, and propose an end-to-end study design which recruits researchers to execute these ideas into full projects, enabling us to study whether these novelty and feasibility judgements result in meaningful differences in research outcome.
From: Chenglei Si [view email].
