Abstract：Large scale galaxy surveys have brought us unprecedented opportunities in exploring various mysteries of our Universe, e.g., dark matter and dark energy, which may lead to exciting leaps in our understanding of fundamental physics. Weak lensing is a direct probe of the large scale structure, as only gravitational physics is involved. We show that to achieve robust constraints on cosmology, weak lensing measurements in both the underdense and overdense regions, as well as shape measurement itself, can be done in untraditional ways to reduce their model-dependences. In the first part of my talk, I show a new way of defining the underdense regions, and how to use their background lensing signals to constrain the equation-of-state of dark energy. Secondly, regarding the halo mass function measurement with weak lensing, I discuss a way of avoiding the Eddington bias and possibly many other sources of errors. Finally, I report the status of our Fourier_Quad shear measurement pipeline, which is accurate, fast, and model-independent.