SlimSAM: 0.1% Data Makes Segment Anything Slim
Current approaches for compressing the Segment Anything Model (SAM) yield commendable results, yet necessitate extensive data to train a new network from scratch. Employing conventional pruning techniques can remarkably reduce data requirements but would suffer from a degradation in performance. To address this challenging trade-off, we introduce SlimSAM, a novel data-efficient SAM compression method that achieves superior performance with extremely less training data. The essence of SlimSAM is encapsulated in the alternate slimming framework which effectively enhances knowledge inheritance under severely limited training data availability and exceptional pruning ratio. Diverging from prior techniques, our framework progressively compresses the model by alternately pruning and distilling distinct, decoupled sub-structures. Disturbed Taylor pruning is also proposed to address the misalignment between the pruning objective and training target, thereby boosting the post-distillation after pruning. SlimSAM yields significant performance improvements while demanding over 10 times less training data than any other existing compression methods. Even when compared to the original SAM, SlimSAM achieves approaching performance while reducing parameter counts to merely 1.4% (9.1M), MACs to 0.8% (23G), and requiring only 0.1% (10k) of the SAM training data. The code is available at http://github.com/czg1225/SlimSAM.
Code (2)
Methods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
SLiMe: Segment Like Me
Significant strides have been made using large vision-language models, like Stable Diffusion (SD), for a variety of downstream tasks, including image editing, image correspondence, and 3D shape generation. Inspired by th…
3D Shape GenerationSegmentationSlimSeg: Slimmable Semantic Segmentation with Boundary Supervision
Accurate semantic segmentation models typically require significant computational resources, inhibiting their use in practical applications. Recent works rely on well-crafted lightweight models to achieve fast inference.…
Knowledge DistillationSegmentationSemantic SegmentationSegment anything model (SAM) for brain extraction in fMRI studies
Brain extraction and removal of skull artifacts from magnetic resonance images (MRI) is an important preprocessing step in neuroimaging analysis. There are many tools developed to handle human fMRI images, which could in…
Brain SegmentationSegmentationSlim-CNN: A Light-Weight CNN for Face Attribute Prediction
We introduce a computationally-efficient CNN micro-architecture Slim Module to design a lightweight deep neural network Slim-Net for face attribute prediction. Slim Modules are constructed by assembling depthwise separab…
AttributeSegment Anything Model for Grain Characterization in Hard Drive Design
Development of new materials in hard drive designs requires characterization of nanoscale materials through grain segmentation. The high-throughput quickly changing research environment makes zero-shot generalization an …
Zero-shot Generalization