Spatiality-guided Transformer for 3D Dense Captioning on Point Clouds
This work addresses 3D dense captioning for enhanced scene understanding in robotics or AR/VR, but it is incremental as it builds on existing transformer architectures with spatial guidance.
The paper tackles the problem of generating natural language descriptions for objects in 3D point clouds, proposing a transformer-based method that incorporates spatial relations, and achieves improvements of 4.94% and 9.61% in CIDEr@0.5IoU over a baseline on two benchmark datasets.
Dense captioning in 3D point clouds is an emerging vision-and-language task involving object-level 3D scene understanding. Apart from coarse semantic class prediction and bounding box regression as in traditional 3D object detection, 3D dense captioning aims at producing a further and finer instance-level label of natural language description on visual appearance and spatial relations for each scene object of interest. To detect and describe objects in a scene, following the spirit of neural machine translation, we propose a transformer-based encoder-decoder architecture, namely SpaCap3D, to transform objects into descriptions, where we especially investigate the relative spatiality of objects in 3D scenes and design a spatiality-guided encoder via a token-to-token spatial relation learning objective and an object-centric decoder for precise and spatiality-enhanced object caption generation. Evaluated on two benchmark datasets, ScanRefer and ReferIt3D, our proposed SpaCap3D outperforms the baseline method Scan2Cap by 4.94% and 9.61% in CIDEr@0.5IoU, respectively. Our project page with source code and supplementary files is available at https://SpaCap3D.github.io/ .