Search
June 18, Tsinghua News Network: In the realm of machine and artificial intelligence perception and interaction, three-dimensional (3D) vision holds paramount importance. Despite significant advancements, current mainstream 3D imaging technologies—including structured light, time-of-flight, and binocular/multi-view stereo vision—are hampered by challenges such as bulky size, high costs, substantial power consumption, and limited resolution of 3D point clouds. Monocular 3D vision, while garnering considerable attention, offers potential relief from these issues but often falters in weakly textured scenes, complicating the reconstruction of intricate 3D shapes. Moreover, existing 3D imaging solutions grapple with scenarios involving near-transparency or high specular reflections.
-
C114 Communication Network -
Communication Home
