用视觉基础模型消除无源目标检测中的源域偏差，提升跨域泛化能力。
Towards Unbiased Source-Free Object Detection via Vision Foundation Models
- 引入视觉基础模型融合特征，通过统一注入与领域感知加权缓解源域偏倚。
- 在多种跨域任务中达到最高精度，合成到真实场景下达61.4% AP。
- 提出轻量级蒸馏版本，适合计算资源受限场景使用。
无源目标检测（SFOD）近年来受到广泛关注，因其无需源域数据即可完成跨域任务，但现有方法存在源域偏差问题，即适应后的模型仍偏向源域，导致泛化性能差且自训练中误差累积。为此，我们提出去偏无源目标检测（DSOD），一种借助视觉基础模型（VFM）的新框架，可有效缓解源域偏差。具体地，提出统一特征注入（UFI）模块，通过简单扩展与领域感知自适应加权将VFM特征融入CNN主干；提出语义感知特征正则化（SAFR），约束特征学习以防止对源域特征过拟合。此外，针对计算受限场景，提出无需VFM的简化版本DSOD-distill，采用新型双教师蒸馏策略。在多个基准测试上广泛实验表明，DSOD显著优于当前最优的SFOD方法，在正常到雾天天气迁移中达48.1% AP，跨场景迁移中达39.3% AP，合成到真实迁移中达61.4% AP。
原文摘要 · Abstract (English)
Source-Free Object Detection (SFOD) has garnered much attention in recent years by eliminating the need of source-domain data in cross-domain tasks, but existing SFOD methods suffer from the Source Bias problem, i.e. the adapted model remains skewed towards the source domain, leading to poor generalization and error accumulation during self-training. To overcome this challenge, we propose Debiased Source-free Object Detection (DSOD), a novel VFM-assisted SFOD framework that can effectively mitigate source bias with the help of powerful VFMs. Specifically, we propose Unified Feature Injection (UFI) module that integrates VFM features into the CNN backbone through Simple-Scale Extension (SSE) and Domain-aware Adaptive Weighting (DAAW). Then, we propose Semantic-aware Feature Regularization (SAFR) that constrains feature learning to prevent overfitting to source domain characteristics. Furthermore, we propose a VFM-free variant, termed DSOD-distill for computation-restricted scenarios through a novel Dual-Teacher distillation scheme. Extensive experiments on multiple benchmarks demonstrate that DSOD outperforms state-of-the-art SFOD methods, achieving 48.1% AP on Normal-to-Foggy weather adaptation, 39.3% AP on Cross-scene adaptation, and 61.4% AP on Synthetic-to-Real adaptation.
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。