Physics
Training algorithm breaks barriers to deep physical neural networks
Research
EPFL
EPFL researchers have developed an algorithm to train an analog neural network just as accurately as a digital one, enabling the development of more efficient alternatives to power-hungry deep learning hardware.
Enhanced AI tracks neurons in moving animals
Research
EPFL
EPFL and Harvard scientists have developed a pioneering new method that uses deep learning and a new technique called ‘targeted augmentation’ to track neurons in moving and deforming animals.
Watching electrons at work
Research
ETH Zurich Empa
Researchers from ETH Zurich, Empa and Stanford have taken snapshots of the crystal structure of perovskite nanocrystals as it was deformed by excited electrons. To their surprise, the deformation straightened out the skewed crystal structure rather than making it more disordered.
In search of life
Research
ETH Zurich
Sascha Quanz searches for traces of life on extrasolar planets orbiting alien stars. The astrophysicist believes that gaining an understanding of life and its origin on Earth is an important stepping stone.
Measuring earthquakes and tsunamis with fibre-optic networks
Research
ETH Zurich
Geophysicists at ETH Zurich have shown that every single wave of a magnitude 3.9 earthquake registers in the noise suppression system of fibre-optic networks. This method can be used to set up close-meshed earthquake and tsunami early warning systems at low cost.
Strange magnetic material could make computing energy-efficient
Research
EPFL
A research collaboration co-led by EPFL has uncovered a surprising magnetic property of an exotic material that might lead to computers that need less than one-millionth of the energy required to switch a single bit.
Page 5 of 23