Wearable sensor data fusion for human stress estimation
Offer N°: SL-DRT-16-0883
Stress is a normal reaction to external constraints. A reaction too much important or lasting too long can lead to physical or mental disorders. Measuring stress level of a person wearing sensors could help to detect too high or long stress levels and act consequently in order to prevent the emergence of associated troubles. Recent works show that it is possible to link non-invasive physiological measures with several stress levels induced in laboratory conditions.
Proposal of a UNUM arithmetic calculation unit and its application to scientific computing
Offer N°: SL-DRT-16-0568
UNUM is a very recent representation (2010) for floating point numbers, which outperforms the classical IEEE 754 representation for accuracy, and accelerates convergence for iterative calculations.
Channel-aware learning and adaption techniques for Deploy & Forget Ultra Low Power Radiofrequency transceivers
Offer N°: SL-DRT-16-0537
In order to considerable increase the lifetime of autonomous communicating nodes for the Internet of Things, it is necessary to find ways to decrease their average power consumption. One
Multimodal lensless microscopy for biomedical applications
Offer N°: SL-DRT-16-0512
The development of miniaturized and high performance CMOS-sensors allowed the emergence of new imaging methods, called “on-chip” microscopy. These methods, generally based on shadow imaging or in-line holography, use simplified and robust instrumentation, and allow to address applications in "point of care" or in resource limited or hostile environments.