E2SAM: A Pipeline for Efficiently Extending SAM’s Capability on Cross-Modality Data via Knowledge Inheritance
Code release for "E2SAM" (BMVC 2023). The brief introduction of our proposed method is shown as:
Abstract: Segment Anything Model (SAM) has achieved brilliant results on many segmentation datasets due to its strong segmentation capability with visual-grouping perception. However, the limitation of the three-channel input means that it is difficult to apply directly to cross-modality data.
Therefore, this paper proposes a pipeline called
- mmengine>=2.0
- pytorch>=2.0
-
Download from SHIFT Dataset, SFDD Dataset (public soon).
SHIFT Dataset Images (1 fps) Train/Val RGB Image Front 12.6G / 2.0G Semantic Segmentation Front 3.9G / 0.667G Depth Maps Front 68.4G / 11.4G -
Training:
python runner_shift_4_sam.py
-
Test:
python infer.py
If you find this paper useful in your research, please consider citing:
Thanks for your attention! If you have any suggestion or question, you can leave a message here or contact us directly: