A diffraction-limited 30-meters class telescope theoretically provides a 10 mas resolution limit in the near infrared. Modern coronagraphs offer the means to take full advantage of this angular resolution allowing to explore at high contrast, the innermost parts of nearby planetary systems to within a fraction of an astronomical unit: an unprecedented capability that will revolutionize our understanding of planet formation and evolution across the habitable zone. A precursor of such a system is the Subaru Coronagraphic Extreme AO project. SCExAO  uses advanced coronagraphic technique for high contrast imaging of exoplanets and disks as close as 1 λ/D from the host star. In addition to unusual optics, achieving high contrast at this small angular separation requires a wavefront sensing and control architecture which is optimized for exquisite control and calibration of low order aberrations. To complement the current near-IR wavefront control system driving a single MEMS type deformable mirror mounted on a tip-tilt mount, two high order and high sensitivity visible wavefront sensors have been integrated to SCEXAO: - a non-modulated Pyramid wavefront sensor (CHEOPS) which is a sensitivity improvement over modulated Pyramid systems now used in high performance astronomical AO, - a non-linear wavefront sensor  designed in 2012 by Subaru Telescope with the collaboration of the NRC-CNRC which is expected to improve significantly the achieved sensitivity of low order aberations measurements. We will present the CHEOPS last results measured in laboratory and during its first light downstream the Subaru AO188 instrument, and then conclude introducing the primary prototype of the SCExAO non-linear curvature wavefront sensor which is planned to be tested on sky in 2014.