Accuracy Threshhold not working

[jo2]

I have a problem that the Accuracy Threshhold is not working in my evaluation. I configured this:

cfg.MODEL.RETINANET.SCORE_THRESH_TEST = 0.5 
 cfg.MODEL.ROI_HEADS.SCORE_THRESH_TEST = 0.5
 cfg.MODEL.PANOPTIC_FPN.COMBINE.INSTANCES_CONFIDENCE_THRESH = 0.5

but no threshhold is applied during the visualization of my test data.

This is how a visualize my test data:

img = cv2.imread(d["file_name"])
outputs = predictor(img)
visualizer = Visualizer(img[:, :, ::-1], metadata=can_metadata, scale=0.8, instance_mode=ColorMode.SEGMENTATION)
vis = visualizer.draw_instance_predictions(outputs["instances"].to("cpu"))
cv2.imwrite("/volume/processed/" + d["file_name"][36:], vis.get_image()[:, :, ::-1])

Do you know a way to solve this issue?

[arnaud-nt2i]

@jo2 did you solved your pb ?

[jo2]

We didn't get the filter to work. Our main problem seemed to be that we did not use a pre trained image resulting in low confidence.

[lamhoangtung]

@jo2 @arnaud-nt2i You can set the threshold properly like this:

    THRESH_TEST = 0.7
    cfg.MODEL.ROI_HEADS.SCORE_THRESH_TEST = THRESH_TEST
    cfg.MODEL.RETINANET.SCORE_THRESH_TEST = THRESH_TEST
    cfg.MODEL.PANOPTIC_FPN.COMBINE.INSTANCES_CONFIDENCE_THRESH = THRESH_TEST
    cfg.MODEL.FCOS.INFERENCE_TH_TEST = THRESH_TEST

[jo2]

I don't know if we tried the PANOPTIC_FPN config but we tried the others and it did not work. Sadly I can't reproduce it anymore because I don't have the code and data any more.

[li-fz]

did you solved the problem,I meet the problem that I can't get the mask ,but it's a square @jo2 @youngwanLEE

[jo2]

No I couldn't solve the problem as I don't have access to the data any more