标题： SAM-PTx：基于文本引导的SAM参数高效并行文本适配器微调 显示英文标题

标题： SAM-PTx: Text-Guided Fine-Tuning of SAM with Parameter-Efficient, Parallel-Text Adapters

摘要： 段分割任何模型（SAM）在基于提示的分割中展示了令人印象深刻的整体性能。 然而，与传统的空间提示（如点和框）相比，语义文本提示的潜力仍未得到充分探索。 本文介绍了SAM-PTx，这是一种利用冻结的CLIP派生文本嵌入作为类级语义指导的参数高效方法，用于适应SAM。 具体而言，我们提出了一种轻量级适配器设计，称为Parallel-Text，它将文本嵌入注入到SAM的图像编码器中，在保持大部分原始架构冻结的同时，实现语义引导的分割。 我们的适配器仅修改每个变压器块的MLP-parallel分支，保留注意力路径以进行空间推理。 通过在COD10K数据集以及COCO和ADE20K的低数据子集上的监督实验和消融研究，我们证明将固定文本嵌入作为输入可以提高分割性能，优于纯粹的空间提示基线。 据我们所知，这是首次在COD10K数据集上使用文本提示进行分割的工作。 这些结果表明，将语义条件整合到SAM的架构中，为以最小计算复杂度进行高效适应提供了一条实用且可扩展的路径。 显示英文摘要

摘要： The Segment Anything Model (SAM) has demonstrated impressive generalization in prompt-based segmentation. Yet, the potential of semantic text prompts remains underexplored compared to traditional spatial prompts like points and boxes. This paper introduces SAM-PTx, a parameter-efficient approach for adapting SAM using frozen CLIP-derived text embeddings as class-level semantic guidance. Specifically, we propose a lightweight adapter design called Parallel-Text that injects text embeddings into SAM's image encoder, enabling semantics-guided segmentation while keeping most of the original architecture frozen. Our adapter modifies only the MLP-parallel branch of each transformer block, preserving the attention pathway for spatial reasoning. Through supervised experiments and ablations on the COD10K dataset as well as low-data subsets of COCO and ADE20K, we show that incorporating fixed text embeddings as input improves segmentation performance over purely spatial prompt baselines. To our knowledge, this is the first work to use text prompts for segmentation on the COD10K dataset. These results suggest that integrating semantic conditioning into SAM's architecture offers a practical and scalable path for efficient adaptation with minimal computational complexity.