Camera-based noise removal
At DxO Labs we have developed a sensor-calibration-based method to provide our noise removal engine with the best possible noise model. The correction algorithm built from this hardware-model-based correction is specifically tailored by DxO Labs for each camera (and more and more cameras are always being profiled) and for all of its capture parameters (ISO speed, Ev bias, color temperature, exposure time, and so on).
One of the targets used by DxO Labs to characterize the camera noise.