.png&w=640&q=75)
SINet: Semantic Segmentation
SINet is an efficient model for precise salient instance segmentation in complex scenes. It innovatively integrates multi-scale feature pyramids with attention mechanisms, employing a dual-branch architecture to process edge details and semantic information separately, while enhancing target localization through interactive feature fusion. With lightweight design, it maintains high accuracy while reducing computational costs, enabling real-time applications. Tested on DUTS and ECSSD datasets, SINet outperforms existing models with fewer parameters, achieving optimal speed-accuracy balance. Its applications span autonomous driving, medical image analysis, and intelligent photo editing.
Source model
- Input shape: 1x3x224x224
- Number of parameters: 91.9Kb
- Model size: 0.4M
- Output shape: 1x2x224x224
The source model can be found here
Performance Reference
Please search model by model name in Model Farm
Inference & Model Conversion
Please search model by model name in Model Farm
License
Source Model: MIT
Deployable Model: APLUX-MODEL-FARM-LICENSE