CLIP-Guided Vision-Language Pre-training for Question Answering in 3D Scenes
This work addresses the challenge of 3D world reasoning for AI systems, representing an incremental advance in vision-language pre-training for 3D domains.
The paper tackles the problem of applying 2D vision-language knowledge to 3D scene understanding by proposing a pre-training method that aligns 3D point cloud features with CLIP embeddings, resulting in state-of-the-art performance on 3D Visual Question Answering.
Training models to apply linguistic knowledge and visual concepts from 2D images to 3D world understanding is a promising direction that researchers have only recently started to explore. In this work, we design a novel 3D pre-training Vision-Language method that helps a model learn semantically meaningful and transferable 3D scene point cloud representations. We inject the representational power of the popular CLIP model into our 3D encoder by aligning the encoded 3D scene features with the corresponding 2D image and text embeddings produced by CLIP. To assess our model's 3D world reasoning capability, we evaluate it on the downstream task of 3D Visual Question Answering. Experimental quantitative and qualitative results show that our pre-training method outperforms state-of-the-art works in this task and leads to an interpretable representation of 3D scene features.