Lightning Talk: Deploying AI Solutions on the Edge with Containers
summary
Describes a concrete pipeline that packages ONNX Runtime, OpenCV and GStreamer into prebuilt containers, pushes them to Azure Container Registry and deploys with Azure IoT Edge so the same inference and business logic runs across GPU, FPGA and other accelerators via ONNX execution providers. Demonstrates a working vision gateway that ingests USB or RTSP streams, runs optimized ONNX models in‑container and emits an RTSP output with inference overlays, removing the need to rewrite apps per device.
AI-written summary · about
chapters (6)
- 0:00Start
- 0:33Frameworks, Toolchain, Al Accelerators
- 1:55ONNX Runtime
- 3:06Container with Abstraction Layer
- 3:46Sample Vision Al Solution with Container
- 4:34Conclusions
related talks
45mFrom Data to AI on The Edge
36mApplied AI on the edge
49mEnable IoT with Edge Computing and Machine Learning
01h 02mAI on the Edge
31mBuild Your Intelligent Edge
30mScalable ML acceleration with ONNX Runtime
43mIntersection of Awesome: Building AI enabled IoT solutions powered by the cloud
07mLightning Talk: AI at the Edge: ONNX Inference in WASM on Featherweight K0s