How to write instance segmentation example result into a coco dataset?

[NowLoadY]

I'm new, I read the document, I simply changed the last sentence of the sample code:

# bproc.renderer.enable_segmentation_output(map_by=["category_id", "instance", "name"])
bproc.renderer.enable_segmentation_output(map_by=["instance"])
# render the whole pipeline
data = bproc.renderer.render()
# write the data to a .hdf5 container
#bproc.writer.write_hdf5(args.output_dir, data)
# Try write the data to coco format
bproc.writer.write_coco_annotations(args.output_dir,
                                    instance_segmaps=data['instance_segmaps'],
                                    instance_attribute_maps=None,
                                    colors=data['colors'],
                                    color_file_format='PNG',
                                    mask_encoding_format='rle',
                                    supercategory='coco_annotations',
                                    append_to_existing_output=True,
                                    segmap_output_key='segmap',
                                    segcolormap_output_key='segcolormap',
                                    rgb_output_key='colors',
                                    jpg_quality=95,
                                    label_mapping=None,
                                    file_prefix='')

I don't know how to deal with

instance_segmaps=data['instance_segmaps'],
instance_attribute_maps=None,
colors=data['colors'],
color_file_format='PNG',
mask_encoding_format='rle',
supercategory='coco_annotations',
append_to_existing_output=True,
segmap_output_key='segmap',
segcolormap_output_key='segcolormap',
rgb_output_key='colors',
jpg_quality=95,
label_mapping=None,
file_prefix='')

the result was not what I expected.
I got this:

Error: Python: Traceback (most recent call last):
  File "<string>", line 1, in <module>
  File "D:\coding\python_learning\BlenderProc-main\examples/basics/semantic_segmentation/main.py", line 44, in <module>
    bproc.writer.write_coco_annotations(args.output_dir,
  File "C:\Users\NowLoadY\.conda\envs\blenderproc\lib\site-packages\blenderproc\python\writer\CocoWriterUtility.py", line 87, in write_coco_annotations     
    raise RuntimeError(f"There is no output registered with key {segcolormap_output_key}. Are you sure you "
RuntimeError: There is no output registered with key segcolormap. Are you sure you ran the SegMapRenderer module with 'map_by' set to 'instance' before?

Could anyone please tell me what to do, I would be very grateful.💌

[NowLoadY]

Solved, instead of using:

# activate depth rendering
bproc.renderer.enable_depth_output(activate_antialiasing=False)
# enable segmentation masks (per class and per instance)
bproc.renderer.enable_segmentation_output(map_by=["category_id", "instance", "name"])

I should use:

# activate normal rendering
bproc.renderer.enable_normals_output()
# enable segmentation masks (per class and per instance)
bproc.renderer.enable_segmentation_output(map_by=["category_id", "instance", "name"])

😀
and refer to examples/advanced/main.py:

# Write data to coco file
bproc.writer.write_coco_annotations(args.output_dir,
                                    instance_segmaps=data['instance_segmaps'],
                                    instance_attribute_maps=data["instance_attribute_maps"],
                                    colors=data['colors'],
                                    color_file_format="JPEG",
                                    segmap_output_key='segmap',
                                    segcolormap_output_key='segcolormap',
                                    rgb_output_key='colors',
                                    label_mapping=None)