The Intel® DevCloud containerized marketplace reference samples enables users to seamlessly build and test containerized AI inference workloads on Intel® hardware specialized for deep learning. The containerized refrence samples contain optimized deep-learning models pre-built with the Intel® Distribution of OpenVINO™ toolkit to do the inferencing on Intel® Core™ CPUs i3, i5, i7 and Xeons.
Each sample contains instructions for:
- How It Works?
- Supported runtime customizations
- Building and running on Intel® DevCloud and your local system
Container applications demonstrating inference pipelines with Intel® Distribution of OpenVINO™ toolkit - Inference Engine.
Application | Description |
---|---|
Safety Gear Detection | Use an optimized and pre-trained MobileNet-SSD neural network to detect people and their safety gear from video input. |
People Counter System | Deploy a smart video IoT solution using a person detection model from Intel® Distribution of OpenVINO™ toolkit to detect and counter people in each frame of a video feed. |
Accelerated Object Detection | Accelerate object detection by using asynchronous inferencing and distributing workloads to multiple types of processing units. |
Tiny YOLO V3 Object Detection | Convert a pre-trained DarkNet YOLO V3 model to TensorFLow, then run accelerated inference using OpenVINO™ for object detection. Learn how to fine-tune an application for optimal performance. |
Benchmark Sample | Learn how to use the Intel® DevCloud benchmarking tool to evaluate the performance of your model's synchronous and asynchronous inference. |
Deep Learning Streamer | Learn how to utilize the GStreamer* plug-in to manage complex media analytics pipelines and boost your AI inferencing capabilities. |
Pneumonia Classification | Classify the probability of pneumonia in X-Ray images using a pre-trained neural network and the Intel® Distribution of OpenVINO™ toolkit. |
TensorFlow* container applications with OpenVINO™ toolkit optimizations.
Application | Description |
---|---|
Object Detection | The sample showcases object detection using YoloV3 TensorFlow Model on OpenVINO™ integration with Tensorflow. |
Classification | The sample is to showcase classification of image with inception V3 Tensorflow model using OpenVINO™ integration with Tensorflow. |
PyTorch* container applications with OpenVINO™ toolkit optimizations.
Application | Description |
---|---|
Image Classification | The sample showcases image classification using ResNet-50 PyTorch Model on OpenVINO™ integration with Torch-ORT. |
Sequence Classification | The sample is to showcase sequence classification of text with BERT PyTorch model using OpenVINO™ integration with Torch-ORT. |