无源目标检测(source -free object detection, SFOD)是指在没有源域数据的情况下,使源检测器适应于未标记的目标域数据。大多数SFOD方法使用平均教师(MT)框架遵循相同的自我训练范式,其中学生模型仅由单个教师模型指导。然而,这种范式很容易陷入训练不稳定性问题,即当教师模型由于领域转移而无法控制地崩溃时,学生模型的...
Source-Free domain adaptive Object Detection (SFOD) is a promising strategy for deploying trained detectors to new, unlabeled domains without accessing source data, addressing significant concerns around data privacy and efficiency. Most SFOD methods leverage a Mean-Teacher (MT) self-training paradigm ...
Source-Free domain adaptive Object Detection (SFOD) is a promising strategy for deploying trained detectors to new, unlabeled domains without accessing sou... I Yoon,H Kwon,J Kim,... 被引量: 0发表: 2024年 Reliable hybrid knowledge distillation for multi-source domain adaptive object detection ...
Method Detector Source-free K2C S2C Source Only Faster R-CNN 36.4 33.7 SFOD-Mosaic (SED) [38] Faster R-CNN ✓ 44.6 42.9 LODS [37] Faster R-CNN ✓ 43.9 - IRG [55] Faster R-CNN ✓ 45.7 43.2 A2SFOD [10] Faster R-CNN ✓ 44.9 44.0 RPL [63] Faster R-CNN ✓ 47.8 50.1 ...
Trinh Le Ba Khanh, Huy-Hung Nguyen, Long Hoang Pham, Duong Nguyen-Ngoc Tran and Jae Wook Jeon Official Pytorch implementation ofDynamic Retraining-Updating Mean Teacher for Source-Free Object Detection, ECCV 2024paper. The overview of our DRU method is presented in the following figure. For mor...
Recent research has proposed various solutions for Source-Free Object Detection (SFOD), most being variations of teacher-student architectures with diverse feature alignment, regularization and pseudo-label selection strategies. Our work investigates simpler approaches and their performance compared to more ...
This study focuses on source-free object detection (SFOD), which adapts a source-trained detector to an unlabeled target domain without using labeled source data. Recent advancements in self-training, particularly with the Mean Teacher (MT) framework, show promise for SFOD deployment. However, the...