# DriveSafe: A Framework for Risk Detection and Safety Suggestions in Driving Scenarios

arXiv: https://arxiv.org/abs/2605.16892 · HTML: https://mlsanity.com/paper/2605.16892/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in VLMs for autonomous driving, one of 9 directions in End-to-end autonomous driving.
- Builds on 29 of its 42 references that are in the map.
- First posted on arXiv 16 May 2026.

## Placed in

- End-to-end autonomous driving / [VLMs for autonomous driving](https://mlsanity.com/driving/vlms-for-autonomous-driving/)

## Builds on (most cited first, 29 of 29)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [Improved Baselines with Visual Instruction Tuning](https://mlsanity.com/paper/2310.03744/) (2023)
- [CIDEr: Consensus-based Image Description Evaluation](https://mlsanity.com/paper/1411.5726/) (2014)
- [Depth Anything V2](https://mlsanity.com/paper/2406.09414/) (2024)
- [Token Merging: Your ViT But Faster](https://mlsanity.com/paper/2210.09461/) (2022)
- [SPICE: Semantic Propositional Image Caption Evaluation](https://mlsanity.com/paper/1607.08822/) (2016)
- [LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init Attention](https://mlsanity.com/paper/2303.16199/) (2023)
- [LLaVA-NeXT-Interleave: Tackling Multi-image, Video, and 3D in Large Multimodal Models](https://mlsanity.com/paper/2407.07895/) (2024)
- [DriveLM: Driving with Graph Visual Question Answering](https://mlsanity.com/paper/2312.14150/) (2023)
- [VideoLLaMA 3: Frontier Multimodal Foundation Models for Image and Video Understanding](https://mlsanity.com/paper/2501.13106/) (2025)
- [GPT-Driver: Learning to Drive with GPT](https://mlsanity.com/paper/2310.01415/) (2023)
- [DRAMA: Joint Risk Localization and Captioning in Driving](https://mlsanity.com/paper/2209.10767/) (2022)
- [Holistic Autonomous Driving Understanding by Bird's-Eye-View Injected Multi-Modal Large Models](https://mlsanity.com/paper/2401.00988/) (2024)
- [Rank2Tell: A Multimodal Driving Dataset for Joint Importance Ranking and Reasoning](https://mlsanity.com/paper/2309.06597/) (2023)
- [Embodied Understanding of Driving Scenarios](https://mlsanity.com/paper/2403.04593/) (2024)
- [HiLM-D: Enhancing MLLMs with Multi-Scale High-Resolution Details for Autonomous Driving](https://mlsanity.com/paper/2309.05186/) (2023)
- [CLAIR: Evaluating Image Captions with Large Language Models](https://mlsanity.com/paper/2310.12971/) (2023)
- [Reinforcement Learning for Autonomous Driving with Latent State Inference and Spatial-Temporal Relationships](https://mlsanity.com/paper/2011.04251/) (2020)
- [RAIN: Reinforced Hybrid Attention Inference Network for Motion Forecasting](https://mlsanity.com/paper/2108.01316/) (2021)
- [Who Make Drivers Stop? Towards Driver-centric Risk Assessment: Risk Object Identification via Causal Inference](https://mlsanity.com/paper/2003.02425/) (2020)
- [HybridNets: End-to-End Perception Network](https://mlsanity.com/paper/2203.09035/) (2022)
- [V2V-LLM: Vehicle-to-Vehicle Cooperative Autonomous Driving with Multimodal Large Language Models](https://mlsanity.com/paper/2502.09980/) (2025)
- [Goal-oriented Object Importance Estimation in On-road Driving Videos](https://mlsanity.com/paper/1905.02848/) (2019)
- [Interaction Graphs for Object Importance Estimation in On-road Driving Videos](https://mlsanity.com/paper/2003.06045/) (2020)
- [RoadSocial: A Diverse VideoQA Dataset and Benchmark for Road Event Understanding from Social Video Narratives](https://mlsanity.com/paper/2503.21459/) (2025)
- [Video Token Sparsification for Efficient Multimodal LLMs in Autonomous Driving](https://mlsanity.com/paper/2409.11182/) (2024)
- [MLLM-SUL: Multimodal Large Language Model for Semantic Scene Understanding and Localization in Traffic Scenarios](https://mlsanity.com/paper/2412.19406/) (2024)
- [Hints of Prompt: Enhancing Visual Representation for Multimodal LLMs in Autonomous Driving](https://mlsanity.com/paper/2411.13076/) (2024)

## Abstract (arXiv)

Comprehensive situational awareness is essential for autonomous vehicles operating in safety-critical environments, as it enables the identification and mitigation of potential risks. Although recent Multimodal Large Language Models (MLLMs) have shown promise on general vision-language tasks, our findings indicate that zero-shot MLLMs still underperform compared to domain-specific methods in fine-grained, spatially grounded risk assessment. To address this gap, we propose DriveSafe, a framework for risk-aware scene understanding that leverages structured natural language descriptions. Specifically, our method first generates spatially grounded captions enriched with multimodal context, including motion, spatial, and depth cues. These captions are then used for downstream risk assessment, explicitly identifying hazardous objects, their locations, and the unsafe behaviors they imply, followed by actionable safety suggestions. To further improve performance, we employ caption-risk pairings to fine-tune a lightweight adapter module, efficiently injecting domain-specific knowledge into the base LLM. By conditioning risk assessment on explicit language-based scene representations, DriveSafe achieves significant gains over both zero-shot MLLMs and prior domain-specific baselines. Exhaustive experiments on the DRAMA benchmark demonstrate state-of-the-art performance, while ablation studies validate the effectiveness of our key design choices. Project page: https://cvit.iiit.ac.in/ research/projects/cvit-projects/drivesafe