Eval json does not match Coco results

[HoopsMcann]

Here is the coco results from running sahi evaluate:

Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=500 ] = 0.506 Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=500 ] = 0.780 Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=500 ] = 0.531 Average Precision (AP) @[ IoU=0.50 | area= small | maxDets=500 ] = 0.723 Average Precision (AP) @[ IoU=0.50 | area=medium | maxDets=500 ] = 0.594 Average Precision (AP) @[ IoU=0.50 | area= large | maxDets=500 ] = 0.890 Average Precision (AP) @[ IoU=0.50:0.95 | area= small | maxDets=500 ] = 0.357 Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=500 ] = 0.295 Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets=500 ] = 0.627 Average Recall (AR) @[ IoU=0.50:0.95 | area= small | maxDets=500 ] = 0.439 Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=500 ] = 0.439 Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets=500 ] = 0.723 COCO evaluation results are successfully exported to vision/doc_inference/eval.json

And here are the results that are written to the eval.json:
{ "bbox_mAP": 0.506, "bbox_mAP50": 0.78, "bbox_mAP75": 0.531, "bbox_mAP_s": 0.357, "bbox_mAP_m": 0.295, "bbox_mAP_l": 0.627, "bbox_mAP50_s": 0.723, "bbox_mAP50_m": 0.594, "bbox_mAP50_l": 0.89, "bbox_mAP_copypaste": "0.506 0.531 0.780 0.357 0.295 0.627 0.723 0.594 0.890" }

All the keys in the eval.json file show precision metrics, but no recall metrics.

[fcakyon]

Hello @HoopsMcann, currently recall metrics are not exported into eval json, you can open a PR if you need that feature 👍