Qualcomm® AI HubAI Hub

Semantic Segmentation Ubuntu

On‑device semantic segmentation with TFLite

A sample app for Ubuntu demonstrating semantic segmentation on a live camera stream via TFLite Runtime. It uses GStreamer for input video and provides a minimal web interface to stream the color‑overlaid output frames, suitable for embedded applications.

View App READMEDiscuss on Slack
1. Choose a compatible model
2. Get source code and dependencies
3. Build and run app

Compatible Models

DeepLabV3-Plus-MobileNet

Deep Convolutional Neural Network model for semantic segmentation.

DeepLabXception

Deep Convolutional Neural Network model for semantic segmentation.

FCN-ResNet50

Fully-convolutional network model for image segmentation.

Information

Operating System

Ubuntu 24.04+

Language

Python

Runtime

TFLite

Use Case

Semantic Segmentation

Join Us on Slack

Join our growing community of 2000+ developers in the AI Hub Slack Channel building the future of AI
Join Now