README.md

Mask R-CNN FP32 inference

Description

This document has instructions for running Mask R-CNN FP32 inference using Intel-optimized TensorFlow.

Datasets and Pretrained Model

Download the MS COCO 2014 dataset. Set the DATASET_DIR to point to this directory when running Mask R-CNN.

Quick Start Scripts

Script name	Description
fp32_inference.sh	Runs inference with batch size 1 using Coco dataset and pretrained model

Run the model

Setup your environment using the instructions below, depending on if you are using AI Kit:

Setup using AI Kit	Setup without AI Kit
AI Kit does not currently support TF 1.15.2 models	To run without AI Kit you will need: Python 3.6 or 3.7 intel-tensorflow==1.15.2 numactl git wget IPython[all] pillow>=8.1.2 cython h5py==2.10.0 imgaug keras==2.0.8 matplotlib numpy==1.16.3 opencv-python pycocotools scikit-image scipy==1.2.0 A clone of the Model Zoo repo git clone https://github.com/IntelAI/models.git

Running Mask R-CNN also requires a clone and particular SHA of the Mask R-CNN model repository. Set the MODEL_SRC_DIR env var to the path of your clone.

git clone https://github.com/matterport/Mask_RCNN.git
cd Mask_RCNN
git checkout 3deaec5d902d16e1daf56b62d5971d428dc920bc
export MODEL_SRC_DIR=$(pwd)

Download pre-trained COCO weights mask_rcnn_coco.h5) from the Mask R-CNN repository release page, and place it in the MaskRCNN directory.

wget -q https://github.com/matterport/Mask_RCNN/releases/download/v2.0/mask_rcnn_coco.h5
cd ..

After your environment is setup, set environment variables to the DATASET_DIR and an OUTPUT_DIR where log files will be written. Ensure that you already have the MODEL_SRC_DIR path set from the previous commands. Once the environment variables are all set, you can run the quickstart script.

# cd to your model zoo directory
cd models

export DATASET_DIR=<path to the dataset>
export OUTPUT_DIR=<directory where log files will be written>
export MODEL_SRC_DIR=<path to the Mask RCNN models repo with pre-trained model>

./quickstart/image_segmentation/tensorflow/maskrcnn/inference/cpu/fp32/fp32_inference.sh

Additional Resources

• To run more advanced use cases, see the instructions here for calling the launch_benchmark.py script directly.
• To run the model using docker, please see the oneContainer workload container:
  https://software.intel.com/content/www/us/en/develop/articles/containers/mask-rcnn-fp32-inference-tensorflow-container.html.