Vous êtes ici


Agrégateur de flux

New approach may open up speech recognition to more languages

ScienceDaily's AI News - mer, 07/12/2016 - 14:07
Researchers have developed a new approach to training speech-recognition systems that doesn't depend on transcription. Instead, their system analyzes correspondences between images and spoken descriptions of those images, as captured in a large collection of audio recordings. The system then learns which acoustic features of the recordings correlate with which image characteristics.

Wall-jumping robot is most vertically agile ever built

ScienceDaily's AI News - mar, 06/12/2016 - 14:27
Roboticists have designed a small robot that can leap into the air and then spring off a wall, or perform multiple vertical jumps in a row, resulting in the highest robotic vertical jumping agility ever recorded.

Birds flying through laser light reveal faults in flight research, study shows

ScienceDaily's AI News - mar, 06/12/2016 - 11:17
Parrotlets flying through a field of lasers and microparticles helped test three popular models that predict the lift generated by flying animals. The work could help develop better flying robots.

No peeking: Humans play computer game using only direct brain stimulation

ScienceDaily's AI News - mar, 06/12/2016 - 10:35
Researchers have published the first demonstration of humans playing a simple, two-dimensional computer game using only input from direct brain stimulation -- without relying on any usual sensory cues from sight, hearing or touch.

Construction of practical quantum computers radically simplified

ScienceDaily's AI News - ven, 02/12/2016 - 10:34
Scientists have invented a ground-breaking new method that puts the construction of large-scale quantum computers within reach of current technology.

Computer learns to recognize sounds by watching video

ScienceDaily's AI News - jeu, 01/12/2016 - 12:10
In recent years, computers have gotten remarkably good at recognizing speech and images: Think of the dictation software on most cellphones, or the algorithms that automatically identify people in photos posted to Facebook. But recognition of natural sounds has lagged behind. That's because most automated recognition systems, whether they process audio or visual information, are the result of machine learning, in which computers search for patterns in huge compendia of training data, say investigators.

What makes Bach sound like Bach? New dataset teaches algorithms classical music

ScienceDaily's AI News - mer, 30/11/2016 - 14:10
MusicNet is the first publicly available large-scale classical music dataset designed to allow machine learning algorithms to tackle a wide range of open challenges -- from automated music transcription to listening recommendations based on the structure of music itself.

Toddler robots help solve how children learn

ScienceDaily's AI News - lun, 28/11/2016 - 11:14
Children may learn new words using the same method as robots. New research suggests that early learning is based not on conscious thought but on an automatic ability to associate objects, which enables babies to quickly make sense of their environment.

Liquid silicon: Multi-duty computer chips could bridge the gap between computation and storage

ScienceDaily's AI News - jeu, 17/11/2016 - 13:46
Computer chips in development could make future computers more efficient and powerful by combining tasks usually kept separate by design, report investigators.

Video Games: When lag time gets you shot

ScienceDaily's AI News - jeu, 17/11/2016 - 11:51
Do you always get 'shot' playing computer games? Do you, like a sore loser, also blame the equipment? Actually, rather surprisingly, you might be justified.

Tech would use drones and insect biobots to map disaster areas

ScienceDaily's AI News - jeu, 17/11/2016 - 10:44
Researchers have developed a combination of software and hardware that will allow them to use unmanned aerial vehicles (UAVs) and insect cyborgs, or biobots, to map large, unfamiliar areas -- such as collapsed buildings after a disaster.

New AI algorithm taught by humans learns beyond its training

ScienceDaily's AI News - mer, 16/11/2016 - 10:22
Researchers have designed an algorithm that learns directly from human instructions, rather than an existing set of examples, and outperformed conventional methods of training neural networks by 160 per cent. But more surprisingly, their algorithm also outperformed its own training by nine per cent -- it learned to recognize hair in pictures with greater reliability than that enabled by the training, marking a significant leap forward for artificial intelligence.

Researchers question if banning of 'killer robots' actually will stop robots from killing

ScienceDaily's AI News - mar, 15/11/2016 - 15:53
A research team has published a paper that implies that the rush to ban and demonize autonomous weapons or "killer robots" may be a temporary solution, but the actual problem is that society is entering into a situation where systems like these have and will become possible.

Real men don't say 'cute'

ScienceDaily's AI News - mar, 15/11/2016 - 12:12
What's in a tweet? From gender to education, the words used on social media carry impressions to others. Using publicly available tweets, social psychologists and computer scientists are helping us to parse out the stereotypes formed by word choices on the social media channel Twitter. Utilizing natural language processing (NLP), a form of artificial intelligence, the researchers show where stereotyping goes from "plausible" to wrong.

Paralyzed ALS patient operates speech computer with her mind

ScienceDaily's AI News - lun, 14/11/2016 - 11:29
A brain implant has been placed in a patient enabling her to operate a speech computer with her mind. The researchers and the patient worked intensively to get the settings right. She can now communicate at home with her family and caregivers via the implant. That a patient can use this technique at home is unique in the world.

Insect vector feeding recognized by machine learning

ScienceDaily's AI News - jeu, 10/11/2016 - 15:32
Scientists have used machine learning algorithms to teach computers to recognize the insect feeding patterns involved in pathogen transmission. The study also uncovers plant traits that might lead to the disruption of pathogen transmission and enable advances in agriculture, livestock and human health.

Artificial-intelligence system surfs web to improve its performance

ScienceDaily's AI News - jeu, 10/11/2016 - 12:02
Of the vast wealth of information unlocked by the Internet, most is plain text. The data necessary to answer myriad questions -- about, say, the correlations between the industrial use of certain chemicals and incidents of disease, or between patterns of news coverage and voter-poll results -- may all be online. But extracting it from plain text and organizing it for quantitative analysis may be prohibitively time consuming.

Dyke inspection robot with an innovative powertrain

ScienceDaily's AI News - mer, 09/11/2016 - 09:03
Future robots that continuously inspect our dykes, don’t come across an electrical charging station every few hours. Using a smart gear box for the robot, a researcher manages to drastically reduce the energy consumption. The energy-autonomous robot comes closer.

New U.S. robotics roadmap calls for increased regulations, education and research

ScienceDaily's AI News - mer, 02/11/2016 - 13:04
A new U.S. Robotics Roadmap calls for better policy frameworks to safely integrate new technologies, such as self-driving cars and commercial drones, into everyday life. The document also advocates for increased research efforts in the field of human-robot interaction to develop intelligent machines that will empower people to stay in their homes as they age. It calls for increased education efforts in the STEM fields from elementary school to adult learners

For they know not what they do: Study on the robotization of office, service professions

ScienceDaily's AI News - mer, 02/11/2016 - 08:00
Robots, common in manufacturing jobs, increasingly spread to office professions. Humanoid or human-like robots already perform tasks in hotels, in stores, and in restaurants. They cook, serve, or advise customers. They communicate like humans via speech, gestures, and sometimes even facial expressions. A real robot hype can be observed in many occupations that are at risk of being replaced by robots/in the process of robotization, say authors of a new report.