COCO People dataset
This dataset is a subset of the original COCO dataset. All annotations were filtered to retain only those corresponding to the person class, effectively removing all other object categories. The complete set of images from the original dataset is preserved — including those that no longer have any associated annotations after filtering. Images without any person annotations are treated as negative samples, helping improve model robustness by providing background-only examples during training.
Object Detection Benchmark (100 epochs)
COCO Metrics - RGB - (640x640) | ONNX
| Model | mAP@0.5 | mAP@0.5..0.95 |
|---|---|---|
| modelpack-csp19-nano-640x640-rgb | 0.480 | 0.223 |
| ultralytics-yolov8n-640x640-rgb | 0.620 | 0.410 |
Timing Benchmark
Time information is not included in this validation because they can change depending on the hardware quality of the cloud servers in EdgeFirst Studio. The timing benchmarks are provided when the models are run on specific platforms such as the i.MX 8M Plus.
COCO Metrics - RGB - (640x640) | TFLite | INT8
| Model | mAP@0.5 | mAP@0.5..0.95 | Time (ms) |
|---|---|---|---|
| modelpack-csp19-nano-640x640-rgb | 0.196 | 0.073 | 10.63 |
| ultralytics-yolov8n-640x640-rgb | 0.583 | 0.374 | 137.43 |
BSP Version
i.MX 8M Plus is flashed with NXP Yocto BSP 6.12.
Dataset Information
Groups:
- train: 64115 Images
- val: 2693 images
It contains 123,287 images in total and one class.
Ungrouped Images
There are 56,479 images in this dataset that are not associated to the train or val groups.
Image Gallery
License
TBA