Detect objects with centroids
FOMO (Faster Objects, More Objects) is a brand-new approach to run object detection models on constrained devices. FOMO is a ground-breaking algorithm that brings real-time object detection, tracking and counting to microcontrollers for the first time. FOMO is 30x faster than MobileNet SSD and can run in <200K of RAM.
In this tutorial, we will explain how to count cars to estimate parking occupancy using FOMO.
View the finished project, including all data, signal processing and machine learning blocks here: Car Parking Occupancy Detection - FOMO.
Limitations of FOMO
FOMO does not output bounding boxes but will give you the object's location using centroids. Hence the size of the object is not available.
FOMO works better if the objects have a similar size.
Objects shouldn’t be too close to each other, although this can be optimized when increasing the image input resolution.
If you need the size of the objects for your project, head to the default Object detection tutorial.
1. Prerequisites
For this tutorial you'll need a supported device. Either:
MCU targets:
Linux-based devices
If you don't have any of these devices, you can also upload an existing dataset through the Uploader or use your mobile phone to connect your device to Edge Impulse. After this tutorial, you can then deploy your trained machine learning model as a C++ library or as a WebAssembly package and run it on your device.
2. Building a dataset
Capturing data
You can collect data from the following devices:
Collecting image data from the Studio - for the Raspberry Pi 4 and the Jetson Nano.
Collecting image data from any of the fully-supported development boards that have a camera.
Alternatively, you can capture your images using another camera, and then upload them directly from the studio by going to Data acquisition and clicking the 'Upload' icon or using Edge Impulse CLI Uploader.
With the data collected we need to label this data. Go to Data acquisition, verify that you see your data, then click on the 'Labeling queue' to start labeling.
Labeling data
Why use bounding box inputs?
To keep the interoperability with other models, your training image input will use bounding boxes although we will output centroids in the inference process. As such FOMO will use in the background translation between bounding boxes and segmentation maps in various parts of the end-to-end flow. This includes comparing sets between the bounding boxes and the segmentation maps to run profiling and scoring.
All our collected images will be staged for annotation at the "labeling queue". Labeling your objects is as easy as dragging a box around the object, and entering a label. However when you have a lot of images, this manual annotation method can become tiresome and time consuming. To make this task even easier, Edge impulse provides 3 AI assisted labeling methods that can help you save time and energy. The AI assisted labeling techniques include:
Using YoloV5 - Useful when your objects are part of the common objects in the COCO dataset.
Using your own trained model - Useful when you already have a trained model with classes similar to your new task.
Using Object tracking - Useful when you have objects that are similar in size and common between images/frames.
For our case, since the 'car' object is part of the COCO dataset, we will use the YoloV5 pretrained model to accelerate this process. To enable this feature, we will first click the Label suggestions dropdown,then select “Classify using YOLOv5.”
From the image above, the YOLOV5 model can already help us annotate more than 90% of the cars without us having to do it manually by our hands.
Rebalancing your dataset
To validate whether a model works well you want to keep some data (typically 20%) aside, and don't use it to build your model, but only to validate the model. This is called the 'test set'. You can switch between your training and test sets with the two buttons above the 'Data collected' widget. If you've collected data on your development board there might be no data in the testing set yet. You can fix this by going to Dashboard > Rebalance dataset.
3. Designing an impulse
One of the beauty of FOMO is its fully convolutional nature, which means that just the ratio is set. Thus, it gives you more flexibility in its usage compared to the classical Object detection method. For this tutorial we have been using 96x96 images but it will accepts other resolutions as long as the images are square.
To configure this, go to Create impulse, set the image width and image height to 96, the 'resize mode' to Fit shortest axis and add the 'Images' and 'Object Detection (Images)' blocks. Then click Save impulse.
Configuring the processing block
As of today, FOMO only accepts Greyscale images. To configure your processing block, click Images in the menu on the left. This will show you the raw data on top of the screen (you can select other files via the drop-down menu), and the results of the processing step on the right. You can use the options to switch between 'RGB' and 'Grayscale' mode, but for now, leave the color depth on 'Greyscale' and click Save parameters.
This will send you to the 'Feature generation' screen. In here you'll:
Resize all the data.
Apply the processing block on all this data.
Create a 3D visualization of your complete dataset.
Click Generate features to start the process.
Afterwards the 'Feature explorer' will load. This is a plot of all the data in your dataset. Because images have a lot of dimensions (here: 96x96x1=9216 features) we run a process called 'dimensionality reduction' on the dataset before visualizing this. Here the 9216 features are compressed down to 3 , and then clustered based on similarity as shown in the feature explorer below.
Configuring the object detection model with FOMO
With all data processed it's time to start training our FOMO model . The model will take an image as input and output objects detected using centroids. For our case, it will show centroids of cars detected on the images.
FOMO is fully compatible with any MobileNetV2 model, and depending on where the model needs to run you can pick a model with a higher or lower alpha. Transfer learning also works (although you need to train your base models specifically with FOMO in mind). Another advantage of FOMO is that it has very few parameters to learn from compared to normal SSD networks making the network even much smaller and faster to train . Together this gives FOMO the capabilities to scale from the smallest microcontrollers all the way to full gateways or GPUs.
To configure FOMO, head over to the ‘Object detection’ section, and select 'Choose a different model' then select one of the FOMO models as shown in the image below.
Make sure to start with a learning rate of 0.001 then click start training. After the model is done you'll see accuracy numbers below the training output. You have now trained your FOMO object detection model!
As you may have noticed from the training results above, FOMO uses F1 Score as its base evaluating metric as compared to SSD MobileNetV2 which uses Mean Average Precision (mAP). Using Mean Average Precision (mAP) as the sole evaluation metric can sometimes give limited insights about a model’ performance. This is particularly true when dealing with datasets with imbalanced classes as it only measures how accurate the predictions are without putting into account how good or bad the model is for each individual class. The combination between F1 score and a confusion matrix gives us both the balance between precision and recall of our model as well as how the model performs for each individual class.
4. Validating your model
With the model trained let's try it out on some test data. When collecting the data we split the data up between a training and a testing dataset. The model was trained only on the training data, and thus we can use the data in the testing dataset to validate how well the model will work in the real world. This will help us ensure the model has not learned to overfit the training data, which is a common occurrence. To validate our model, we will go to Model testing and select Classify all.
Given the little training data we had and the few cycles we trained on, we got an accuracy of 84.62% which can be improved further. To see a classification in detail, we will head to live classification, and select one image from our test sample.click the three dots next to an item, and select Show classification. We can also capture new data directly from your development board from here.
From the test image above, our model was able to detect 16 cars out of the actual possible 18 which is a good performance.
5. Running the model on your device
With the impulse designed, trained and verified you can deploy this model back to your device. This makes the model run without an internet connection, minimizes latency, and runs with minimum power consumption. Edge Impulse can package up the complete impulse - including the preprocessing steps, neural network weights, and classification code - in a single C++ library or model file that you can include in your embedded software.
Running the impulse on a Linux-based device
From the terminal just run edge-impulse-linux-runner. This will build and download your model, and then run it on your development board. If you're on the same network you can get a view of the camera, and the classification results directly from your dev board. You'll see a line like:
Open this URL in a browser to see your impulse running!
Running the impulse on a fully supported MCU
Go to the Deployment tab, on Build firmware section and select the board-compatible firmware to download it.
Follow the instruction provided to flash the firmware to your board and head over to your terminal and run the edge-impulse-run-impulse --debug command:
You'll also see a URL you can use to view the image stream and results in your browser:
Running the impulse using a generated Arduino Library
To run using an Arduino library, go to the studio Deployment tab on Create Library section and select Arduino Library to download your custom Arduino library. Go to your Arduino IDE, then click on Sketch >> Include Library >> Add .Zip ( Your downloaded Arduino library). Make sure to follow the instruction provided on Arduino's Library usage. Open Examples >> Examples from custom library and select your library. Upload the ''Portenta_H7_camera'' sketch to your Portenta then open your serial monitor to view results.
Integrating the model in your own application
Congratulations! You've added object detection using FOMO to your sensors. Now that you've trained your model you can integrate your impulse in the firmware of your own edge device, see Deploy your model as a C++ library or the Edge Impulse for Linux documentation for the Node.js, Python, Go and C++ SDKs that let you do this in a few lines of code and make this model run on any device.
Here's an example of sending a text message through Twilio when an object is seen.
Or if you're interested in more, see our tutorials on Continuous motion recognition or Adding sight to your sensors. If you have a great idea for a different project, that's fine too. Edge Impulse lets you capture data from any sensor, build custom processing blocks to extract features, and you have full flexibility in your Machine Learning pipeline with the learning blocks.
We can't wait to see what you'll build! 🚀
Last updated
