IGLOSS: Image Generation for Lidar Open-vocabulary Semantic Segmentation
This work addresses the problem of semantic segmentation in autonomous driving for lidar data, enabling open-vocabulary capabilities without extensive labeled data, though it is incremental in improving existing methods.
The paper tackles zero-shot open-vocabulary semantic segmentation of 3D lidar data by using image generation from text to create prototypes, circumventing the image-text modality gap in VLMs like CLIP, and achieves state-of-the-art results on nuScenes and SemanticKITTI benchmarks.
This paper presents a new method for the zero-shot open-vocabulary semantic segmentation (OVSS) of 3D automotive lidar data. To circumvent the recognized image-text modality gap that is intrinsic to approaches based on Vision Language Models (VLMs) such as CLIP, our method relies instead on image generation from text, to create prototype images. Given a 3D network distilled from a 2D Vision Foundation Model (VFM), we then label a point cloud by matching 3D point features with 2D image features of these prototypes. Our method is state-of-the-art for OVSS on nuScenes and SemanticKITTI. Code, pre-trained models, and generated images are available at https://github.com/valeoai/IGLOSS.