不微调模型，用推理时融合三种解码策略提升动态4D重建性能。
Gallileo-4D: Frozen Backbone Ensemble for Dynamic 4D Reconstruction
- 冻结预训练4D骨干网络，推理时融合三种解码配置。
- 在无任何梯度更新下获得0.58356 APD，比微调效果更好。
- 适合资源受限但需高精度动态重建的场景。
我们介绍参加PhysAI动态4D重建挑战赛的方案，在最终排行榜上以0.58356 APD位列第3（共27支队伍），且全程未进行任何梯度更新。原计划尝试13种微调配置，但其中12种反而降低评测分数，且11种同时提升了局部验证得分。我们发现此反常现象源于基准数据集结构：仅25%的评估集属于训练阶段发布的数据变体，因此对可用数据过拟合会损害剩余75%数据依赖的预训练特征。为此，系统冻结骨干网络，将全部计算预算用于推理阶段，通过凸加权融合三种解码配置——时间步长为3、水平翻转测试时增强、密集步长为1。该集成方法相较冻结基线提升+0.041 APD，超过所有训练实验，且训练成本为零。
原文摘要 · Abstract (English)
We describe our entry to the PhysAI Dynamic 4D Reconstruction Challenge, which placed third of 27 teams at 0.58356 APD on the final leaderboard, without a single gradient update. This was not the plan: of thirteen fine-tuning configurations of a pre-trained 4D backbone, twelve degraded the challenge score, and eleven of those twelve improved local validation at the same time. We trace this inversion to the structure of the benchmark: only 25% of the evaluation set belongs to the data variant released for training, so updates that fit the available data damage the pre-trained features the remaining 75% relies on. Our system therefore freezes the backbone and spends its budget at inference time, fusing three decoding configurations -- temporal stride-3, horizontal-flip test-time augmentation, and dense stride-1 -- under a convex weighting. The ensemble recovers +0.041 APD over the frozen baseline, more than any training run achieved, at zero training cost.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。