globaledgesoft Goto Github PK

ai-with-gstreamer-on-rb5

Qualcomm provides its own version of GStreamer which allows you to load and execute AI models

aiml-board-app

Contains the code with regards to the steps undertaken for training the model with caffe framework on custom data as well as the converted model (.dlc) file which has been deployed in the Qualcomm Android Application

aiml-dashcam-app

The objective of this repository is to create an android application which would have the deep model trained and converted to the Qualcomm NPE format and thus deployed

ar-based-home-automation-using-snapdragon-spaces-xr

Smart Home Automation by Using AR with XR Interaction

audio-based-chatbot-rb5

This project is intended to build and deploy an interactive communication robot on the RB5 platform. It responds to the set of queries or questions which are being asked to the system by any human being.

autonomousrb5

This project is for the demonstration of Wake-word based Trigger to RB5, Stop Sign Detection, Reinforcement Learning based Lane Tracking & Following, and Obstacle Detection which automates the RB5 to run autonomously on maps.

autoware.ai-on-rb5

Autoware.ai is an Autonomous Vehicle development software stack. This project demonstrate that how to deploy Autoware.ai on Qualcomm Robotics RB5 Platform.

aws-iot-integration-for-object-detection-using-cm2290-on-android

The goal of this project is to develop an AI-powered application that can detect various objects and send the detected data to AWS IoT services. This application will be built for the Android platform and will be integrated with the Thundercomm TurboX CM2290 LA (Linux Android) Development Kit.

aws-iot-on-qcs610

AWS IoT C++/Python SDK integration into the QCS610 and Data posting to AWS IOT console

aws-lex-c610

Integration of AWS Lex with C610

azure-iot-qcs610

Integration of Azure IoT into QCS 610

azureml

Demonstration of Azure ML & its services to use for robotics related or ML based cognitive solutions

barcode-reader-app-c610

Barcode region detection and recognition of the EAN-13 type barcode in the given image

behavior-cloning-with-rosbag-on-rb5

This project is demonstrated to showcase the capabilities of RB5 platform of Qualcomm for performing the ROSBAG recording & Playback.

bringup-of-autonomousroboticsoftwarestack-with-snpe

The project demonstrates the integration of SNPE with Autonomous Driving/Robotics Software Stack on Qualcomm's Robotics Platform RB5 and Object detection node created in Autoware.ai stack is integrated with Aslan, an open-source full-stack software based on ROS framework for self-driving applications.

calibration-of-lidar-camera-for-object-detection-onrb5

The use case demonstrates the Calibration of Camera & Lidar sensor in order to detect the front object from the ISP camera and gives the respective distance for each detected object.

center-stage

collision-avoidance-rb5

This project is for avoiding a collision using lidar sensor & algorithm on RB5.

containerized-pipeline-for-multimodel-inference-onrb5

This project is to demonstrate the multiple container based pipeline for performing the inference on Qualcomm Robotics RB5 Platform.

deeplabv3-application-using-neural-processing-sdk

The application demonstrates the Image Semantic Segmentation considering the input image in portrait mode and changes the background to grayscale except for the persons.

depth-estimation-on-cm2290-with-snpe-on-android

This project is intended to build and deploy a depth estimation without depth sensor to generate depth map from RGB camera frames using monocular depth estimation model PackNet-SfM on CM2290 with SNPE on android Application with Integration on Thundercomm TurboX CM2290 LA (Linux Android) DK (Development Kit).

depth-estimation-without-using-depth-sensor-on-rb5

The goal of this project is to avoid using the depth sensor, which can be heavy in cost, and try to estimate the depth in meters in realtime scenarios only using the RGB color image. The project is implemented on Qualcomm's Robotics Platform RB5.

docker-deployment-on-rb5

This project is indented to demonstrate the deployment of Docker and running the multiple containers on Qualcomm Robotics RB5 Platform.

drivable-area-segmentation-using-hybridnets-on-qualcomm-rb5

The project is used to demonstrate the Drivable area segmentation on Qualcomm’s Robotics Platform RB5. HybridNets is the architecture used to implement this solution on top of the BDD100K Dataset.

edge-ai-security-using-aws-boto3

The goal of this project is to explore security techniques for passing images from edge devices to the cloud using AWS Boto3 object encryption techniques

efficientdet-based-object-detection-on-rb5

This project is intended to build and deploy EfficientDet model integrated Object Detection that can detect various objects and will be integrated with Qualcomm's RB5.

enable-camera-streaming-on-cm2290-using-aws-kinesis-on-android

The goal of this project is to build and deploy a live video camera streaming to AWS Kinesis video streaming service. Implemented Android Application for Thundercomm TurboX CM2290 LA (Linux Android) DK (Development Kit).

face-detection-c610

Demo project detailing the face-detection system on C610

facerecognition-on-cm2290-with-snpe-on-android

This project is intended to build and deploy the Face recognition with SNPE on Android using ArcFace Algorithm. Integrating the same on Thundercomm TurboX CM2290 LA (Linux Android) DK (Development Kit), which helps to deploy AI Camera Applications on edge device.

facial-expression-recognition-using-keras

The application demonstrates the Facial Expression Recognition which works in real-time on Snapdragon mobile platforms