[2603.04165] PlaneCycle: Training-Free 2D-to-3D Lifting of Foundation Models Without Adapters

Computer Science > Computer Vision and Pattern Recognition arXiv:2603.04165 (cs) [Submitted on 4 Mar 2026] Title:PlaneCycle: Training-Free 2D-to-3D Lifting of Foundation Models Without Adapters Authors:Yinghong Yu, Guangyuan Li, Jiancheng Yang View a PDF of the paper titled PlaneCycle: Training-Free 2D-to-3D Lifting of Foundation Models Without Adapters, by Yinghong Yu and 2 other authors View PDF HTML (experimental) Abstract:Large-scale 2D foundation models exhibit strong transferable representations, yet extending them to 3D volumetric data typically requires retraining, adapters, or architectural redesign. We introduce PlaneCycle, a training-free, adapter-free operator for architecture-agnostic 2D-to-3D lifting of foundation models. PlaneCycle reuses the original pretrained 2D backbone by cyclically distributing spatial aggregation across orthogonal HW, DW, and DH planes throughout network depth, enabling progressive 3D fusion while preserving pretrained inductive biases. The method introduces no additional parameters and is applicable to arbitrary 2D networks. Using pretrained DINOv3 models, we evaluate PlaneCycle on six 3D classification and three 3D segmentation benchmarks. Without any training, the lifted models exhibit intrinsic 3D fusion capability and, under linear probing, outperform slice-wise 2D baselines and strong 3D counterparts, approaching the performance of fully trained models. With full fine-tuning, PlaneCycle matches standard 3D architectures, highlig...