AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

MV-STRIDE: Enabling MLLMs to Master Multi-View Spatial Reasoning via Hierarchical Capability Modeling

Jin Xu, Xiaojian Huang, zhang zhihong, Luo Zhuodong, Xuejin Chen, Jie Zhao, Xin Liu, Wang Xinzhi, Jiansheng Wei

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Spatial reasoning in VLMs · 0 citations (OpenAlex)

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 changedWhereMeasuredEffect
add Stage 1 foundational SFT before cold start and RL (ablation)MV-STRIDE-RL · MMSI-Benchaccuracy: 29.20 -> 37.50Improved
add Stage 2 CoT cold start before RL (ablation)MV-STRIDE-RL · CV-Benchaccuracy: 69.37 -> 85.94Improved
add Stage 2 CoT cold start before RL (ablation)MV-STRIDE-RL · 3DSRBenchaccuracy: 43.16 -> 59.43Improved
add Stage 3 RL after cold start (ablation)MV-STRIDE · MMSI-Benchaccuracy: 36.70 -> 37.50Improved
add Stage 3 RL after cold start (ablation)MV-STRIDE · 3DSRBenchaccuracy: 61.08 -> 59.43Hurt
switch from short-answer SFT (Stage 1) to CoT cold start (Stage 2) (ablation)MV-STRIDE · MMSI-Benchaccuracy: 37.20 -> 36.70Hurt
use all-in-one SFT instead of multi-stage training with RLMV-STRIDE · MMSI-Benchaccuracy: 38.90 -> 37.50Hurt
combine synthetic and real-world Level III data instead of synthetic only (ablation)MV-STRIDE SFT · MMSI-Benchaccuracy: 36.70 -> 39.40Improved
use real-world instead of synthetic data only (ablation)MV-STRIDE SFT · MMSI-Benchaccuracy: 36.70 -> 37.50Improved
add Level I single-view perception data (ablation)MV-STRIDE Stage 1 · ViewSpatial-Benchaccuracy: 48.77 -> 51.17Improved
add Level II data on top of Level III only (ablation)MV-STRIDE Stage 1 · MMSI-Benchaccuracy: 36.80 -> 37.30Improved
fine-tune base model with MV-STRIDE data (SFT Full)Qwen3-VL-8B · MMSI-Benchaccuracy: 29.20 -> 38.90Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy