JetSeg: Efficient Real-Time Semantic Segmentation Model for Low-Power GPU-Embedded Systems

TL;DR

JetSeg is a novel, efficient real-time semantic segmentation model optimized for low-power GPU-embedded systems, combining lightweight architecture, innovative convolution strategies, and a new loss function to outperform existing models in speed and parameter efficiency.

Contribution

The paper introduces JetSeg, a new semantic segmentation model with a lightweight encoder, novel convolution blocks, and a combined loss function, optimized for embedded systems and achieving superior speed and efficiency.

Findings

JetSeg reduces parameters by 46.70M compared to state-of-the-art models.

JetSeg is up to 2x faster on NVIDIA Titan RTX and Jetson Xavier.

JetSeg outperforms existing models in speed and parameter efficiency.

Abstract

Real-time semantic segmentation is a challenging task that requires high-accuracy models with low-inference times. Implementing these models on embedded systems is limited by hardware capability and memory usage, which produces bottlenecks. We propose an efficient model for real-time semantic segmentation called JetSeg, consisting of an encoder called JetNet, and an improved RegSeg decoder. The JetNet is designed for GPU-Embedded Systems and includes two main components: a new light-weight efficient block called JetBlock, that reduces the number of parameters minimizing memory usage and inference time without sacrificing accuracy; a new strategy that involves the combination of asymmetric and non-asymmetric convolutions with depthwise-dilated convolutions called JetConv, a channel shuffle operation, light-weight activation functions, and a convenient number of group convolutions for embedded systems, and an innovative loss function named JetLoss, which integrates the Precision, Recall, and IoUB losses to improve semantic segmentation and reduce computational complexity. Experiments demonstrate that JetSeg is much faster on workstation devices and more suitable for Low-Power GPU-Embedded Systems than existing state-of-the-art models for real-time semantic segmentation. Our approach outperforms state-of-the-art real-time encoder-decoder models by reducing 46.70M parameters and 5.14% GFLOPs, which makes JetSeg up to 2x faster on the NVIDIA Titan RTX GPU and the Jetson Xavier than other models. The JetSeg code is available at https://github.com/mmontielpz/jetseg.