LEDNet
A lightweight encoder-decoder neural network for real-time semantic segmentation on resource-constrained devices.
What is LEDNet?
LEDNet is a lightweight convolutional neural network designed for real-time semantic segmentation, a computer vision task that assigns a class label to each pixel in an image. It solves the problem of high computational cost in dense estimation tasks, enabling efficient deployment on mobile and embedded devices. The model uses an asymmetric encoder-decoder architecture with channel split/shuffle operations and an attention pyramid network to achieve high speed and accuracy.
Researchers and engineers working on real-time computer vision applications, particularly those deploying semantic segmentation models on resource-constrained hardware like mobile devices or edge computing platforms.
Developers choose LEDNet for its exceptional balance of speed and accuracy with minimal parameters (under 1M). It outperforms other lightweight models on standard benchmarks like Cityscapes while maintaining real-time inference speeds, making it ideal for production systems with latency constraints.
Overview
LEDNet: A Lightweight Encoder-Decoder Network for Real-time Semantic Segmentation
Use Cases
Best For
- Real-time semantic segmentation on mobile or embedded devices
Related Projects
Found a gem we're missing?
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
