Available for request or viewing on ResearchGate .
While based on YOLOv8, this "LS" (Lightweight and Scalable) variant is highly cited for its use of Multi-Scale Ghost Convolution (MSGConv) and efficiency gains of up to 55% FPS. Full Text Access: View the full paper on ResearchGate . Key Technical Features of LS-Models
In these "Lightweight" (LS) models, the following components are typically highlighted in the full papers: Ls Models (10) mp4
This paper introduces a lightweight model designed for underwater vehicles, utilizing Region Scaling (RS) loss and self-attention mechanisms to improve small-object detection in complex environments.
Integrated into the neck or head of the network to capture global context without the heavy computational cost of standard transformers. Available for request or viewing on ResearchGate
Reduces parameters and FLOPs while maintaining feature extraction quality.
Based on your search for "LS Models (10)", there are several recent publications that match this technical profile: Key Technical Features of LS-Models In these "Lightweight"
Replaces standard loss functions to better handle small or multi-scale objects.