Publications (6)
Can the Query-based Object Detector Be Designed with Fewer Stages?
Jialin Li, Weifu Fu, Yuhuan Lin +2
Query-based object detectors have made significant advancements since the publication of DETR. However, most existing methods still rely on multi-stage encoders and decoders, or a…
IIDM: Inter and Intra-domain Mixing for Semi-supervised Domain Adaptation in Semantic Segmentation
Weifu Fu, Qiang Nie, Jialin Li +6
Despite recent advances in semantic segmentation, an inevitable challenge is the performance degradation caused by the domain shift in real applications. Current dominant approach…
YOLO-PEFT: Parameter-Efficient Fine-Tuning on YOLO Family
Xu Lin, WenJie Nie, Jinlong Peng +4
Generic parameter-efficient fine-tuning (PEFT) methods transferred from language models can fail silently on real-time detectors, whose heterogeneous operators and detection-specif…
PET-DINO: Unifying Visual Cues into Grounding DINO with Prompt-Enriched Training
Weifu Fu, Jinyang Li, Bin-Bin Gao +6
Open-Set Object Detection (OSOD) enables recognition of novel categories beyond fixed classes but faces challenges in aligning text representations with complex visual concepts and…
Decision Boundary-aware Knowledge Consolidation Generates Better Instance-Incremental Learner
Qiang Nie, Weifu Fu, Yuhuan Lin +5
Instance-incremental learning (IIL) focuses on learning continually with data of the same classes. Compared to class-incremental learning (CIL), the IIL is seldom explored because…
LORS: Low-rank Residual Structure for Parameter-Efficient Network Stacking
Jialin Li, Qiang Nie, Weifu Fu +4
Deep learning models, particularly those based on transformers, often employ numerous stacked structures, which possess identical architectures and perform similar functions. While…