Search papers, labs, and topics across Lattice.
Semantic-Fast-SAM (SFS) is introduced as a semantic segmentation framework leveraging FastSAM, a CNN-based reimplementation of SAM, and a Semantic-Segment-Anything (SSA) labeling strategy. This approach aims to achieve real-time performance without compromising accuracy by assigning meaningful categories to the masks generated by FastSAM. SFS achieves comparable accuracy to SAM-based methods on Cityscapes and ADE20K while being ~20x faster, and demonstrates strong open-vocabulary segmentation capabilities using CLIP-based semantic heads.
Get real-time semantic segmentation with "segment-anything" capabilities, as Semantic-Fast-SAM matches SAM's accuracy at 1/20th the inference time.
We propose Semantic-Fast-SAM (SFS), a semantic segmentation framework that combines the Fast Segment Anything model with a semantic labeling pipeline to achieve real-time performance without sacrificing accuracy. FastSAM is an efficient CNN-based re-implementation of the Segment Anything Model (SAM) that runs much faster than the original transformer-based SAM. Building upon FastSAM's rapid mask generation, we integrate a Semantic-Segment-Anything (SSA) labeling strategy to assign meaningful categories to each mask. The resulting SFS model produces high-quality semantic segmentation maps at a fraction of the computational cost and memory footprint of the original SAM-based approach. Experiments on Cityscapes and ADE20K benchmarks demonstrate that SFS matches the accuracy of prior SAM-based methods (mIoU $\approx 70.33$ on Cityscapes, 48.01 on ADE20K) while achieving $\sim 20 \times$ inference time than SSA. We also show that SFS effectively handles open-vocabulary segmentation by leveraging CLIP-based semantic heads, outperforming recent open-vocabulary models on broad class labeling. This work enables practical real-time semantic segmentation with the “segmentanything” capability, broadening the applicability of foundation segmentation models in robotics scenarios. The implementation is available at https://github.com/KBH00/Semantic-Fast-SAM.
