Wyniki wyszukiwania:

Although research in computer vision for recognizing 3D objects in photographs dates back to the 1960s, progress was relatively slow until the turn of the millennium, and only now do we see the emergence of effective techniques for recognizing object categories with different appearances under large variations in the observation conditions. Tremendous progress has been achieved in the past five years, thanks largely to the integration of new data representations, such as invariant semi-local features, developed in the computer vision community with the effective models of data distribution...