Object detection
Object detection takes an image and outputs information about the class and number of objects, position, (and, eventually, size) in the image.
Edge Impulse provides several object detection model architecture options:
- Using FOMO
- Using MobileNetV2
| Specifications | MobileNetV2 SSD FPN | FOMO |
|---|---|---|
| Labelling method | Bounding boxes | Bounding Boxes |
| Input size | 320x320 | Square (any size) |
| Image format | RGB | Greyscale & RGB |
| Output | Bounding boxes | Centroids |
| MCU | ❌ | ✅ |
| CPU/GPU | ✅ | ✅ |
| Limitations | (1) Works best with big objects. (2) Models use high compute resources (in the edge computing world). (3) Image size is fixed. | (1) Works best when objects have similar sizes and shapes. (2) The size of the objects are not available. (3) Objects should not be too close to each other |