Fork from google-coral/examples-camera

This repo contains a fork of Google-coral's exemple projects from which i have deviated a program which identifies faces.
The scope of the app is to track how many pull-ups does each team member do.
The camera is mounted on a pull-up bar. So it will see only the faces of those who did a full pullup.

Setup

Hardware - the case

I have printed this default case.
And made a system to be able to change the angle of the camera and its height.

Internal Env

Needed to build and install OpenCV in order to save to video. Also check out this guy's comment.
Make sure to have the latest edgetpu version. atm

Implementation

1. Used Gstreamer to receive images.
2. Used Face detector to identify faces on each received image.
3. If any face is found, start a video session, until X (5-for now) seconds have passed without seeing any face.
4. Count the face is found <-> face is not found transitions in order to get the count of all pullups.
5. Use Face detector's output to crop faces and classify between team members.
   Classification was implemented using Coral's weight imprinting. Checkout the bash script from imprinting_classification/imprint.sh for more details.
6. Save the classified team memeber's name, date, number of pullups, and evidence location (the video) to a csv.

Improvements/TODOs:

1. Use OpenCV to trigger face detection only when movement is detected.
2. Switch from face classification using classification imprinting of weights to face recognition using embedding vectors' distance. This, in theory, would make it more flexible to adding more people.
3. All paths are hard coded. Make them relative.

-- Google Coral's readme --

Edge TPU simple camera examples

This repo contains a collection of examples that use camera streams together with the Edge TPU Python API.

Installation

Before you start using the examples run the download_models.sh script in order to download a selection of models. These canned models will be downloaded and extracted to a new folder all_models.

Further requirements may be needed by the different camera libraries, check the README file for the respective subfolder.

Contents

• Gstreamer Python examples using gstreamer to obtain camera images. These examples work on Linux using a webcam, Raspberry Pi with the Raspicam and on the Coral DevBoard using the Coral camera. For the former two you will also need a Coral USB Accelerator to run the models.
• Raspicam Python example using picamera. This is only intended for Raspberry Pi and will require a Coral USB Accelerator. Use install_requirements.sh to make sure all the dependencies are present.
• PyGame Python example using pygame to obtain camera frames. Use install_requirements.sh to make sure all the dependencies are present.
• OpenCV Python example using OpenCV to obtain camera frames. Use install_requirements.sh to make sure all the dependencies are present.

Canned models

For all the demos in this repository you can change the model and the labels file by using the flags flags --model and --labels. Be sure to use the models labeled _edgetpu, as those are compiled for the accelerator - otherwise the model will run on the CPU and be much slower.

For classification you need to select one of the classification models and its corresponding labels file:

inception_v1_224_quant_edgetpu.tflite, imagenet_labels.txt
inception_v2_224_quant_edgetpu.tflite, imagenet_labels.txt
inception_v3_299_quant_edgetpu.tflite, imagenet_labels.txt
inception_v4_299_quant_edgetpu.tflite, imagenet_labels.txt
mobilenet_v1_1.0_224_quant_edgetpu.tflite, imagenet_labels.txt
mobilenet_v2_1.0_224_quant_edgetpu.tflite, imagenet_labels.txt

mobilenet_v2_1.0_224_inat_bird_quant_edgetpu.tflite, inat_bird_labels.txt
mobilenet_v2_1.0_224_inat_insect_quant_edgetpu.tflite, inat_insect_labels.txt
mobilenet_v2_1.0_224_inat_plant_quant_edgetpu.tflite, inat_plant_labels.txt

For detection you need to select one of the SSD detection models and its corresponding labels file:

mobilenet_ssd_v1_coco_quant_postprocess_edgetpu.tflite, coco_labels.txt
mobilenet_ssd_v2_coco_quant_postprocess_edgetpu.tflite, coco_labels.txt
mobilenet_ssd_v2_face_quant_postprocess_edgetpu.tflite, coco_labels.txt