Mask Rcnn inference result through trt.engine does not have segmentation points

Accelerated Computing Intelligent Video Analytics TAO Toolkit
1

The output of Mask Rcnn inference through trt.engine is a json file. But I do not understand the segmentation field. How can I have mask array or polygon points?

        "segmentation": {
            "counts": "TUm<2db03@1S^O3NJWa0\\1i^OcN17[`0b1a_OeNX`0o1N2M3M4M4L4J6M2M2O0K6L3O101O2N2N001N2O2N1O1O1O1O1O1N2N1O1O2O100O1O0O2O0O10000000001O000000000O100O11O001O000000001O0O0100000000000000000O100O1O1O1O2O00002M10O01N1N3O2M2O2K4000O10O2O1N3K6H5O01O01O1O2N8H2N1O2M2O1O2M4M5dMh_Ol1l`Z3",
            "size": [
                600,
                1000
            ]
        }
2

There is no update from you for a period, assuming this is not an issue anymore.
Hence we are closing this topic. If need further support, please open a new one.
Thanks

See MaskRCNN — TAO Toolkit 3.22.05 documentation, the mask_rcnn inference can have output annotated image directory.

3

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.