Dynamic Resolution Routing for Efficient Egocentric Grounding
2026-08-03 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors study how to help AI systems better find small objects in video recorded from a first-person view, which needs very clear detail. They note that current methods that try to speed this up by cutting down on the number of image pieces the AI looks at don't reliably focus on important objects. To fix this, they propose SmartRes, a method that first looks at a blurry full image and then zooms in only on important parts with high detail using a smart controller. Their experiments show it can work faster and almost as accurately as using the full detailed image, especially for spotting small objects.
Egocentric visual groundingMultimodal Large Language ModelsVisual token processingToken reductionDynamic resolution routingForeground-background imbalanceRouterEgo4DObject-centric regionsInference speed
Authors
Huixin Sun, Wangbo Zhao, Fanyue Wei, Qiuxia Lin, Pengzhan Sun, Angela Yao
Abstract
Egocentric visual grounding requires high-resolution inputs to localize small objects. However, scaling Multimodal Large Language Models to this domain is constrained by the excessive cost of visual token processing. We identify that current efficient strategies based on token reduction are unreliable for selecting object-centric spatial evidence. To overcome this, we propose SmartRes, a framework that performs efficiency optimization in the pixel space via dynamic resolution routing. SmartRes first encodes a low-resolution view for global context and uses a lightweight router to activate high-resolution patches in object-centric regions and constructs an order-preserving visual sequence. To further enable robust routing under severe foreground-background imbalance, we introduce a margin-regularized routing objective that increases foreground-background logit separation and improves foreground recall. Experiments on Ego4D and EgoIntention show that SmartRes reduces visual tokens by up to 67% while retaining 86.4% of full-resolution performance, and achieves up to 1.66X faster inference than state-of-the-art token reduction methods with higher accuracy. Furthermore, strong performance on small object grounding indicates the effectiveness of SmartRes towards egocentric applications. Code will be publicly available.