No pred_mask in the predictor output

[vishakhalall]

I am using this to evaluate a trained model on some test images

from detectron2.config import get_cfg
from detectron2.utils.visualizer import Visualizer, ColorMode
from detectron2.engine import DefaultTrainer, DefaultPredictor

cfg.MODEL.WEIGHTS = f"model_final.pth" 
cfg.MODEL.ROI_HEADS.SCORE_THRESH_TEST = 0.75
predictor = DefaultPredictor(cfg)

test_dir = "DetectronDataset/val/images"
for im_p in os.listdir(test_dir):
    im = cv2.imread(f"{test_dir}/{im_p}")
    outputs = predictor(im)  
    v = Visualizer(im,metadata=test_metadata,instance_mode=ColorMode.IMAGE)
    out = v.draw_instance_predictions(outputs["instances"].to("cpu"))
    img = out.get_image()
    print(outputs["instances"])
    cv2.imwrite(f'{output_dir}/{im_p}', out.get_image())

However I noticed that the images do not have the segmentation masks. I tried printing the output of the model and got the following

Instances(num_instances=5, image_height=1280, image_width=720, fields=[pred_boxes: Boxes(tensor([[153.3457, 113.8107, 453.8591, 560.3398],
        [ 77.2884, 298.6120, 624.6890, 489.1781],
        [253.5901,  62.7516, 565.4692, 625.5353],
        [367.0150, 172.0205, 574.5746, 541.7480],
        [ 61.4621,  25.3327, 696.5031, 629.1113]], device='cuda:0')), scores: tensor([0.8329, 0.8182, 0.8029, 0.7972, 0.7571], device='cuda:0'), pred_classes: tensor([0, 0, 0, 0, 0], device='cuda:0')])

As observed, there is no pred_masks in the output. I'm not sure why.

[github-actions]

You've chosen to report an unexpected problem or bug. Unless you already know the root cause of it, please include details about it by filling the issue template.
The following information is missing: "Instructions To Reproduce the Issue and Full Logs"; "Your Environment";

[github-actions]

Requested information was not provided in 7 days, so we're closing this issue.

Please open new issue if information becomes available. Otherwise, use github discussions for free-form discussions.