Note: All images in this directory, unless specified otherwise, are licensed under CC BY-NC 4.0.
| Figure number | Description | Notes |
|---|---|---|
| 14-1 | Building an AI cat sprinkler system, like this Havahart Spray Away Motion Detector | |
| 14-2 | Running a familiar photo on Google’s Vision AI API to obtain object detection results | |
| 14-3 | Object detection prediction in the ready-to-use Jupyter Notebook from the TensorFlow Models repository | |
| 14-4 | Running a real-time object detector model on an Android device | |
| 14-5 | Creating a new Object Detection project in CustomVision.ai | |
| 14-6 | Dashboard with bounding box and class name | |
| 14-7 | Measuring improvement in percent mean average precision with increasing number of images per class | |
| 14-8 | Detected Simpsons characters with the final model, represented by US congress members of the same first name (see note at the beginning of this section) | |
| 14-9 | A timeline of different object detection architectures (image source: Recent Advances in Deep Learning for Object Detection by Xiongwei Wu et al.) | Page 3, Figure 2 |
| 14-10 | The effect of object detection architecture as well as the backbone architecture (feature extractor) on the percent mean average precision and prediction time | Page 8, Figure 2 |
| 14-11 | A visual representation of the IoU ratio | |
| 14-12 | IoU illustrated; predictions from better models tend to have heavier overlap with the ground truth, resulting in a higher IoU | |
| 14-13 | Using NMS to find the bounding box that best represents the location of the object in an image | |
| 14-14 | Photographs of objects taken in a variety of different settings to train an object detector model | |
| 14-15 | Some creative photographs taken during the process of building a diverse currency dataset | |
| 14-16 | Click the Open Dir button and then select the directory that contains the training data | |
| 14-17 | Select the Create RectBox from the panel on the left to make a bounding box that covers the cat | |
| 14-18 | Each image is accompanied by an XML file that contains the label information and the bounding box information | |
| 14-19 | Colorizing hair with ModiFace app by accurately mapping the pixels belonging to the hair | |
| 14-20 | Image segmentation performed on frames from a dashcam (CamVid dataset) | |
| 14-21 | Detected objects along with their classes from the SmartDeviceBox refrigerator | |
| 14-22 | An aerial photograph of wildebeest taken from a small survey airplane | |
| 14-23 | The 2013 Kumbh Mela, as captured by an attendee | Image credit: Seba Della y Sole Bossio, used under CC BY 2.0 |
| 14-24 | Face detection feature on Seeing AI | |
| 14-25 | Detecting traffic lights and signs on a self-driving car using the NVIDIA Drive Platform (image source) |
