CoVeR: Coverage-Based Token Pruning for Multi-View 3D Reasoning in VLMs
CoVeR, a training-free coverage-based token pruner, preserves 93.5% of VLM 3D-reasoning performance using only about 8% of visual tokens.
Researchers introduce CoVeR, a deterministic, training-free selector that chooses visual tokens to cover every region of a multi-view 3D scene using only token coordinates. Unlike learned-importance and voxelization pruners, it enforces an exact per-scene token budget, avoids saturation plateaus, and prevents near-duplicate selections. Experiments across four vision-language models show it surpasses prior state of the art by 3.9 percentage points on average across three 3D reasoning benchmarks.
- CoVeR is training-free and relies only on token coordinates
- Preserves 93.5% of full-token performance with ~8% of visual tokens
- Beats prior SOTA by 3.9 points on average across three 3D benchmarks
- Works as a plug-and-play module across four VLMs
From the original article
1 min read at huggingface.co
Representing a 3D scene as multi-view images allows 2D VLMs to reason in 3D by reusing priors from pre-training, sidestepping the scarcity of annotated 3D data. However, it produces thousands of redundant visual tokens whose cost grows with every view. Existing visual token pruners fall into two families, each limited in the 3D multi-view setting.
By Nhat-Tan Bui, Varshini Elangovan, Arun Reddy Anugu · huggingface.co
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.