Visual Prompt Based Reasoning for Offroad Mapping using Multimodal LLMs
This work addresses the problem of off-road mapping for autonomous navigation by providing a unified framework that reduces the need for multiple specialized models, though it appears incremental as it builds on existing segmentation and VLM technologies.
The paper tackles off-road autonomy by introducing a zero-shot approach that uses SAM2 for segmentation and a vision-language model to reason about drivable areas, eliminating the need for separate terrain-specific models. It surpasses state-of-the-art trainable models on high-resolution segmentation datasets and enables full-stack navigation in an Isaac Sim offroad environment.
Traditional approaches to off-road autonomy rely on separate models for terrain classification, height estimation, and quantifying slip or slope conditions. Utilizing several models requires training each component separately, having task specific datasets, and fine-tuning. In this work, we present a zero-shot approach leveraging SAM2 for environment segmentation and a vision-language model (VLM) to reason about drivable areas. Our approach involves passing to the VLM both the original image and the segmented image annotated with numeric labels for each mask. The VLM is then prompted to identify which regions, represented by these numeric labels, are drivable. Combined with planning and control modules, this unified framework eliminates the need for explicit terrain-specific models and relies instead on the inherent reasoning capabilities of the VLM. Our approach surpasses state-of-the-art trainable models on high resolution segmentation datasets and enables full stack navigation in our Isaac Sim offroad environment.