Machine Learning on the Edge

on YouTube ↗28m

" summary

Hands-on setup of a Jetson Nano Developer Kit: flash a 16+ GB microSD, add a Raspberry Pi camera, cooling fan and install NVIDIA builds of TensorFlow/PyTorch; heavy training happens in the cloud (PaperSpace) then exported as TensorRT for on-device inference. A live imagenet camera demo runs in real time. A home project pairs the Nano with a NodeMCU collecting PM2.5, CO2 and temp sensors into InfluxDB and plans an LSTM trained in the cloud, inference at the edge.

AI-written summary · about

" chapters (4)

1. 0:00Introduction
2. 0:21Edge computing vs cloud
3. 22:26Project implementation
4. 24:02Conclusion and Q&A

" related talks

• 59mML and the IoT: Living on the EdgeBrandon D Satrom · NDC Oslo 2019
• 26mBuilding a Cloud Native Edge Platform for Accelerating AI InferenceJanakiram MSV · ContainerDays 2022
• 01h 01mML and the IoT: Living on the EdgeBrandon D Satrom · NDC London 2020
• 55mEdgeAI - AI in the Palm of Your HandDon Ward · MidWest DevFest 2020
• 49mEnable IoT with Edge Computing and Machine LearningJared Rhodes · Update Conference Prague 2018
• 55mUsing TensorFlow for Performing Deep Learning at the EdgeAditya Bhattacharya · Developers Summit 2021
• 24mInnovation and Solutions for Embedded and Edge AIBarrie Mullins · Industrial IoT Virtual Summit 2020
• 31mThe Rise of No-Code AI Vision at the EdgeAjeet Singh Raina, Lakshantha Dissanayake · Docker Community All-Hands #6