paper-with-me

Papers

Convolution Meets LoRA: Parameter Efficient Finetuning for Segment Anything Model

2024-01-31 · Zihan Zhong, Zhiqiang Tang, Tong He, Haoyang Fang, Chun Yuan

The Segment Anything Model (SAM) stands as a foundational framework for image segmentation. While it exhibits remarkable zero-shot generalization in typical scenarios, its advantage diminishes when applied to specialized domains like medical imagery and remote sensing. To address this limitation, this paper introduces Conv-LoRA, a simple yet effective parameter-efficient fine-tuning approach. By integrating ultra-lightweight convolutional parameters into Low-Rank Adaptation (LoRA), Conv-LoRA can inject image-related inductive biases into the plain ViT encoder, further reinforcing SAM's local prior assumption. Notably, Conv-LoRA not only preserves SAM's extensive segmentation knowledge but also revives its capacity of learning high-level image semantics, which is constrained by SAM's foreground-background segmentation pretraining. Comprehensive experimentation across diverse benchmarks spanning multiple domains underscores Conv-LoRA's superiority in adapting SAM to real-world semantic segmentation tasks.

📄 PDF Abstract BibTeX arXiv:2401.17868

Code (1)

autogluon/autogluon 공식 구현 pytorch

Tasks

Image Segmentationparameter-efficient fine-tuningSegmentationSemantic SegmentationZero-shot Generalization

Methods 이 논문이 사용한 방법론

SAM 설명 없음

Similar Papers 제목 키워드 기반

A Benchmark Study of Segmentation Models and Adaptation Strategies for Landslide Detection from Satellite Imagery

2026-04-17 · Md Kowsher, Weiwei Zhan, Chen Chen arxiv

Landslide detection from high resolution satellite imagery is a critical task for disaster response and risk assessment, yet the relative effectiveness of modern segmentation architectures and finetuning strategies for t…

parameter-efficient fine-tuning

When Scaling Meets LLM Finetuning: The Effect of Data, Model and Finetuning Method

2024-02-27 · Biao Zhang, Zhongtao Liu, Colin Cherry, Orhan Firat

While large language models (LLMs) often adopt finetuning to unlock their capabilities for downstream applications, our understanding on the inductive biases (especially the scaling properties) of different finetuning me…

Machine Translation

LoRA Meets Dropout under a Unified Framework

2024-02-25 · Sheng Wang, Liheng Chen, Jiyue Jiang, Boyang Xue 외

With the remarkable capabilities, large language models (LLMs) have emerged as essential elements in numerous NLP applications, while parameter-efficient finetuning, especially LoRA, has gained popularity as a lightweigh…

Multi LoRA Meets Vision: Merging multiple adapters to create a multi task model

2024-11-21 · Ege Kesim, Selahattin Serdar Helli

Parameter efficient finetuning (PEFT) methods are widely used in LLMs and generative models in computer vision. Especially one can use multiple of these during inference to change the behavior of the base model. In this …

ConvMAE: Masked Convolution Meets Masked Autoencoders

2022-05-08 · Peng Gao, Teli Ma, Hongsheng Li, Ziyi Lin 외

Vision Transformers (ViT) become widely-adopted architectures for various vision tasks. Masked auto-encoding for feature pretraining and multi-scale hybrid convolution-transformer architectures can further unleash the po…

Computational Efficiencyimage-classificationImage ClassificationObject Detection+1