We present Open3D-VQA, a novel benchmark for evaluating MLLMs' ability to reason about complex spatial relationships from an aerial perspective.The QAs are automatically generated from spatial ...
Abstract: 3D Gaussian Splatting (3DGS) has emerged as a powerful representation for neural scene reconstruction, offering high-quality novel view synthesis while maintaining computational efficiency.
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback