No Captions, No Problem: Captionless 3D-CLIP Alignment with Hard Negatives via CLIP Knowledge and LLMs
This addresses the challenge of multimodal alignment for 3D data in computer vision, though it appears incremental as it builds on existing CLIP and contrastive learning frameworks.
The study tackled the problem of aligning 3D objects with text and images without using captions, by introducing unsupervised methods to mine hard negatives and a contrastive pipeline, resulting in comparable or superior performance in 3D classification and significant improvements in cross-modal retrieval.
In this study, we explore an alternative approach to enhance contrastive text-image-3D alignment in the absence of textual descriptions for 3D objects. We introduce two unsupervised methods, $I2I$ and $(I2L)^2$, which leverage CLIP knowledge about textual and 2D data to compute the neural perceived similarity between two 3D samples. We employ the proposed methods to mine 3D hard negatives, establishing a multimodal contrastive pipeline with hard negative weighting via a custom loss function. We train on different configurations of the proposed hard negative mining approach, and we evaluate the accuracy of our models in 3D classification and on the cross-modal retrieval benchmark, testing image-to-shape and shape-to-image retrieval. Results demonstrate that our approach, even without explicit text alignment, achieves comparable or superior performance on zero-shot and standard 3D classification, while significantly improving both image-to-shape and shape-to-image retrieval compared to previous methods.