Home

Disclaimer

I often use this project, but I saw it abandoned and without a public repository on github. Also, part of the project remained unfinished for a long time. I implemented some of the author's ideas and decided to make the results publicly available.

Faster-COCO-Eval base

This package wraps a facebook C++ implementation of COCO-eval operations found in the pycocotools package. This implementation greatly speeds up the evaluation time for coco's AP metrics, especially when dealing with a high number of instances in an image.

Base Usage

from faster_coco_eval import COCO, COCOeval_faster
....

iouType = "segm"
useCats = False

cocoGt = COCO(prepared_coco_in_dict_or_path_to_coco_file)
cocoDt = cocoGt.loadRes(prepared_anns_or_path_to_anns_file)

cocoEval = COCOeval_faster(cocoGt, cocoDt, iouType, extra_calc=True)
cocoEval.params.maxDets = [len(cocoGt.anns)]

if not useCats:
    cocoEval.params.useCats = 0

cocoEval.evaluate()
cocoEval.accumulate()
cocoEval.summarize()

print(cocoEval.stats_as_dict)

Further use is identical to the official library.

Plot PRE-REC Curve

from faster_coco_eval import COCO
from faster_coco_eval.extra import Curves

cocoGt = COCO(....)
cocoDt = cocoGt.loadRes(....)

cur = Curves(cocoGt, cocoDt, iou_tresh=0.5, iouType='segm')
cur.plot_pre_rec()

Preview images + GT with plotly

image_preview_count = 2
cocoGt = COCO(....)
preview = PreviewResults(cocoGt, iouType="segm")
preview.display_tp_fp_fn(
    data_folder=.....,
    image_ids=list(cocoGt.imgs.keys())[:image_preview_count],
    display_gt=True,
)