AI RESEARCH

Balancing Saliency and Coverage: Semantic Prominence-Aware Budgeting for Visual Token Compression in VLMs

arXiv CS.CV

ArXi:2603.14892v1 Announce Type: new Large Vision-Language Models (VLMs) achieve strong multimodal understanding capabilities by leveraging high-resolution visual inputs, but the resulting large number of visual tokens creates a major computational bottleneck. Recent work mitigates this issue through visual token compression, typically compressing tokens based on saliency, diversity, or a fixed combination of both.