Parameter-efficient fine-tuning for instance segmentation on COCO with transformer backbones

Research and applications in artificial intelligence have recently shifted with the rise of large pretrained models, which deliver state-of-the-art results across numerous tasks. However, the substantial increase in parameters introduces a need for parameter-efficient training strategies. Despite significant advancements, limited research has explored parameter-efficient fine-tuning (PEFT) methods in the context of transformer-based models for instance segmentation. Addressing this gap, this study investigates the effectiveness of PEFT methods, specifically adapters and Low-Rank Adaptation (Lo

Research goal: Do parameter-efficient fine-tuning methods like LoRA maintain instance segmentation performance on COCO when applied to other transformer backbones beyond ViT (e.g., Swin Transformers) compared to full fine-tuning?

Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 8.8/10.