The ssd mobilenet v1 caffe network can be used for object detection and can detect 20 different types of objects (This model was pre-trained with the Pascal VOC dataset). More information can be found at: https://github.com/chuanqi305/MobileNet-SSD. The list of objects that this network can detect are:
aeroplane bicycle bird boat bottle bus car cat chair cow diningtable dog horse motorbike person pottedplant sheep sofa train tvmonitor
This sample utilizes the OpenVINO Inference Engine from the OpenVINO Deep Learning Development Toolkit and was tested with the 2020.1 release.
The provided Makefile does the following
- Downloads the prototxt and caffe weight files using the model downloader from the Open Model Zoo.
- Compiles an IR (Intermediate Representation) for the model.
- Takes an image/camera input, loads the IR file, and runs an inference using the SSD Mobilenet model.
- name: ‘data’, shape: [1x3x300x300], Expected color order is BGR.
- name: ‘detection_out’, shape: [1, 1, N, 7] - where N is the number of detected bounding boxes. For each detection, the description has the format: [image_id, label, conf, x_min, y_min, x_max, y_max].
Note: This sample can also be used with a web cam. To run with a webcam, use the following command:
make run INPUT=cam
Provided Makefile describes various targets that help with the above mentioned tasks.
Runs a sample application with the network.
Shows makefile possible targets and brief descriptions.
Makes the follow items: deps, data.
Uses the network description and the trained weights files to generate an IR (intermediate representation) format file. This file is later loaded on the Neural Compute Stick where the inferences on the network can be executed.
Checks required packages that aren’t installed as part of the OpenVINO installation.
Uninstalls requirements that were installed by the sample program.
Removes all the temporary and target files that are created by the Makefile.