This Lab is based on Prof.Marcelo Rovai chapter on Object Detection in the ML Systems open book.
Lab Goal: This lab will explore FOMO: “Faster Objects, More Objects”, a novel object detection technique designed to work efficiently on microcontrollers. Unlike traditional object detection models that compute bounding boxes, FOMO identifies the locations of multiple objects with centroids, significantly reducing computational complexity and memory usage. This makes it highly suitable for real-time applications on resource-constrained devices, such as in IoT deployments for smart agriculture, low-cost healthcare, and industrial monitoring.
We will now be exploring Object Detection on microcontrollers.
The main task with Image Classification models is to produce a list of the most probable object categories present on an image, for example, to identify a tabby cat just after his dinner:
What happens if there is not a dominant category on the image?
The model identifies the above image completely wrong as an “ashcan,” possibly due to the color tonalities.
The model used in all previous examples is the MobileNet, trained with a large dataset, the ImageNet.
To solve this issue, we need another type of model, where not only multiple categories (or labels) can be found but also where the objects are located on a given image.
As we can imagine, such models are much more complicated and bigger, for example, the MobileNetV2 SSD FPN-Lite 320x320, trained with the COCO dataset. This pre-trained object detection model is designed to locate up to 10 objects within an image, outputting a bounding box for each object detected. The below image is the result of such a model running on a Raspberry Pi:
Those models used for Object detection (such as the MobileNet SSD or YOLO) usually have several MB in size, which is OK for use with Raspberry Pi but unsuitable for use with embedded devices, where the RAM usually is lower than 1M Bytes.