OVIR-3D: Open-Vocabulary 3D Instance Retrieval Without Training on 3D Data
It addresses the challenge of 3D instance retrieval for robotics applications like navigation and manipulation, leveraging more accessible 2D data, but is incremental as it builds on existing 2D methods.
This work tackles the problem of open-vocabulary 3D object instance retrieval by proposing OVIR-3D, a method that achieves this without training on 3D data, using multi-view fusion of text-aligned 2D region proposals, and demonstrates effectiveness in experiments on public datasets and a real robot.
This work presents OVIR-3D, a straightforward yet effective method for open-vocabulary 3D object instance retrieval without using any 3D data for training. Given a language query, the proposed method is able to return a ranked set of 3D object instance segments based on the feature similarity of the instance and the text query. This is achieved by a multi-view fusion of text-aligned 2D region proposals into 3D space, where the 2D region proposal network could leverage 2D datasets, which are more accessible and typically larger than 3D datasets. The proposed fusion process is efficient as it can be performed in real-time for most indoor 3D scenes and does not require additional training in 3D space. Experiments on public datasets and a real robot show the effectiveness of the method and its potential for applications in robot navigation and manipulation.