3 accepted papers
The absolute positioning accuracy of robots is a primary factor limiting their applications. In order to enhance the efficiency and precision of robot calibration, this study introduces a method that utilizes the Levy flight and Sparrow Search Algorithm (LSSA) to optimize the measurement pose of the
"We propose an efficient Video-Language Alignment (ViLA) network. Our ViLA model addresses both efficient frame sampling and effective cross-modal alignment in a unified way. In our ViLA network, we design a new learnable text-guided Frame-Prompter together with a cross-modal distillation (QFormer-D…
We introduce Amazon Berkeley Objects (ABO), a new large-scale dataset designed to help bridge the gap between real and virtual 3D worlds. ABO contains product catalog images, metadata, and artist-created 3D models with complex geometries and physically-based materials that correspond to real, househ…