Research map / Vision-language and multimodal / Multimodal LLMs
DH-VLM: Dual-Horizon Cooperative Latent Reasoning for Autonomous Driving
Ziyi Song, Chen Xia, Hang Yu, Sheng Zhou, Zhisheng Niu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add HLA (hierarchical latent aggregation) instead of MLP (ablation) | DH-VLM · DAIR-V2X | avg L2 error (m): 2.31 -> 1.87 | Improved |
| add HLA instead of MLP (ablation) | DH-VLM · DAIR-V2X | avg collision rate (%): 0.39 -> 0.19 | Improved |
| add IDLE instead of MLP (ablation) | DH-VLM · DAIR-V2X | avg L2 error (m): 2.48 -> 1.87 | Improved |
| add IDLE instead of MLP (ablation) | DH-VLM · DAIR-V2X | avg collision rate (%): 0.34 -> 0.19 | Improved |
| add fusion in infrastructure vision encoder (ablation) | infrastructure VLM · DAIR-V2X | avg L2 error (m): 5.19 -> 3.92 | Improved |
| add fundamental scene understanding QA data (ablation) | infrastructure VLM with fusion · DAIR-V2X | avg L2 error (m): 3.92 -> 2.52 | Improved |
| add ego-personalized comprehension QA data (ablation) | infrastructure VLM with fusion and F QA · DAIR-V2X | avg L2 error (m): 2.52 -> 1.72 | Improved |
| fuse 8 intermediate layers instead of only the final layer (ablation) | DH-VLM (5:1) · DAIR-V2X | avg L2 error (m): 2.06 -> 1.87 | Improved |
| increase fused layers from 8 to 16 (ablation) | DH-VLM (5:1) · DAIR-V2X | avg L2 error (m): 1.87 -> 1.88 | No clear effect |
| reduce fused layers from 8 to 4 (ablation) | DH-VLM (5:1) · DAIR-V2X | avg L2 error (m): 1.87 -> 1.93 | Hurt |
| lower communication frequency from 5:1 to 10:1 (ablation) | DH-VLM (8 layers) · DAIR-V2X | avg L2 error (m): 1.87 -> 2.21 | Hurt |
| raise communication frequency from 5:1 to 1:1 (ablation) | DH-VLM (8 layers) · DAIR-V2X | avg L2 error (m): 1.87 -> 1.80 | Improved |
Closest papers by meaning
- OneVL: One-Step Latent Reasoning and Planning with Vision-Language Explanation (NeurIPS 2026 · similarity 0.89)
- SOLVE: Synergy of Language-Vision and End-to-End Networks for Autonomous Driving (CVPR 2025 · similarity 0.89)
- Drive-R1: Bridging Reasoning and Planning in VLMs for Autonomous Driving with Reinforcement Learning (AAAI 2026 · similarity 0.89)
- LaST-VLA: Thinking in Latent Spatio-Temporal Space for Vision-Language-Action in Autonomous Driving (NeurIPS 2026 · similarity 0.89)
- DriveAgent-R1: Advancing VLM-based Autonomous Driving with Active Perception and Hybrid Thinking (ICLR 2026 · similarity 0.89)
- More Than Meets the Eye? Uncovering the Reasoning-Planning Disconnect in Training Vision-Language Driving Models (NeurIPS 2026 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
