Archive for the ‘Artificial Intelligence’ Category.
February 21, 2012
Video game characters have become increasingly realistic. Novel AIs can do many interesting things, but their behavior is still far different from how actual people act. Academics from the Technical University of Munich have a new project that is seeking to use neural agents in a virtual reality environment. These avatars will be able to perform tasks that most humans can do like moving or jumping. The characters will also exhibit facial expressions as well. Their virtual world has a basic physics engine that is similar to what is in most video games. A main feature of this work is that artificial brains control the avatars. The technology uses a neurosim cluster that emulates the synthetic mind. They are employing NVIDIA GPGPU cards along with software named ANNarchy in order to carry out these functions. This program emulates biologically plausible neural networks. There will also be an immersive way for human users to interact with the digital creations. Putting motions sensors on an actual person’s face and body could help the machine detect changes in behavior. This should allow the avatars to perfect how they react to a subject’s emotions or other situations over time. Empathic computers in augmented reality are a potential outcome. Perhaps cloud severs could send this to portable devices in the future. Emotional processes are sometimes devalued, but they are really an important aspect of intelligence. Dysregulation of these perceptions may cause serious issues for an organism. Feelings are a method that your mind uses to guide what you do in an efficient manner. Read More »
November 10, 2011
The Defense Advanced Research Projects Agency is seeking to develop novel theoretical models to help explain how intelligence arises from complex systems. This will allow them to construct electronic circuitry that mimics neurological functioning. Researchers located at the California NanoSystems Institute in the University of California are working on this undertaking. They are creating what they call self-organized critical (SOC) states of open dissipative nonlinear systems. This is a route to go beyond binary CMOS technology. The system may achieve a wide range of behavior in an unstructured environment due to built in plasticity. It sounds like someone just threw a bunch of interesting buzzwords together to cook up a new idea. Perhaps this will succeed where others have failed. They are looking into four different approaches to accomplish this. These include a ferroelectric cellular lattice, a nickel-copper ferromagnetic system, optical rogue waves and an atomic switch array. Those components appear to be novel and they have not been proven to the same degree as other options. The academics are building hardware-based neural networks, as opposed to ones modeled with software. This is somewhat similar to other neuromorphic attempts. A recent paper published in the journal “Advanced Materials” details their progress on making the atomic switch arrays. Read More »
October 17, 2011
In the brain, biological neural networks are constantly firing away and performing advanced calculations. People are completely oblivious to these operations. Our awareness of different emotional states seems so natural. It can be hard to believe that it is all just the result of complex computational processes. The skull encapsulates one of the most complicated machines imaginable. The Moneta version 2.0 project is developing an artificial mind that can carry out some of the tasks a small animal is capable of doing. Many mammals have an amazing ability to navigate their surroundings without the need to use GPS. For these creatures, it is almost instinctive to be able to maneuver around swiftly. Next generation robots may be more like the Von Neumann probes found in the film *batteries not included. They might scavenge for their energy source and explore new territories by themselves without any human intervention. Read More »
September 22, 2011
A program recently took place to discuss laboratory directed research and development science. The LDRD meeting covered a number of innovations. The main goal is to keep the country competitive in a variety of domains. The meeting is associated with the US’s Los Alamos National Lab. A few of the items are emerging trends that can potentially be disruptive in nature. One technology is an ultra-low field MRI. It only needs micro or MilliTesla strength magnetic fields to image biology. This compares to clinical devices that have a much higher Tesla rating. There is hope that this will power machines that require less superconducting cables to run. The magnets consume a lot of liquid helium to maintain a low temperature. Any way of reducing this would help to lessen expenses. They are combining MRI with magnetoencephalography to obtain the best aspects of each technology. A mixture of the two could boost the temporal and spatial capabilities of scanning neurons (see Probing Brain Dynamics by Ultra-low Field Magnetic Resonance PDF). Read More »
July 21, 2011
The US Air Force envisions a future where autonomous vehicles, brain machine interfaces and augmented humans are deployed by the military. A paper entitled “Technology Horizons” details some of the intriguing possible scenarios. The outline seeks to be more realistic than some futurism about what will be possible in the 2010-2030 time frame. Many different technologies are evolving at a fairly rapid pace and will alter significantly the ways wars are fought. Improved radar capacities, advanced missile seekers, high speed ramjet propulsion systems, directed energy weapons and remote piloted air-crafts are just a few of the research areas that the paper briefly covers. The introduction of intelligent planes is mentioned as another major goal for the year 2018. Read More »
July 21, 2011
HRL, in conjunction with the Defense Advanced Research Projects Agency, is developing computer hardware inspired by the human brain. Future neuromorphic technology could augment the capabilities of autonomous military vehicles and usher in a new era of advanced next generation robotics.
CNES promotes the idea that the neural system in the brain is an example of such a complex adaptive system. A key goal of CNES is to explain how computations in the brain can help explain the realization of complex behaviors such as perception, planning, decision making and navigation due to brain-body-environment interactions.
Read More »
Video game characters have become increasingly realistic. Novel AIs can do many interesting things, but their behavior is still far different from how actual people act. Academics from the Technical University of Munich have a new project that is seeking to use neural agents in a virtual reality environment. These avatars will be able to perform tasks that most humans can do like moving or jumping. The characters will also exhibit facial expressions as well. Their virtual world has a basic physics engine that is similar to what is in most video games. A main feature of this work is that artificial brains control the avatars. The technology uses a neurosim cluster that emulates the synthetic mind. They are employing NVIDIA GPGPU cards along with software named ANNarchy in order to carry out these functions. This program emulates biologically plausible neural networks. There will also be an immersive way for human users to interact with the digital creations. Putting motions sensors on an actual person’s face and body could help the machine detect changes in behavior. This should allow the avatars to perfect how they react to a subject’s emotions or other situations over time. Empathic computers in augmented reality are a potential outcome. Perhaps cloud severs could send this to portable devices in the future. Emotional processes are sometimes devalued, but they are really an important aspect of intelligence. Dysregulation of these perceptions may cause serious issues for an organism. Feelings are a method that your mind uses to guide what you do in an efficient manner. Read More »
