Segmentation annotation

[Reagan1311]

Hi, thanks very much for making the dataset public. I noticed that there are object categoy and bounding box annotations in the xml files, but no segmentation annotaion is provided. May I ask where is the segmentation annotation? Thanks!

<annotation>
  <folder>GAZEBOdataset</folder>
 <filename>0.jpg</filename>
  <size>
    <width>640</width>
    <height>480</height>
    <depth>3</depth>
  </size>
  <object>
    <name>bowl</name>
    <pose>Unspecified</pose>
    <truncated>0</truncated>
    <difficult>0</difficult>
    <bndbox>
      <xmin>208</xmin>
      <ymin>240</ymin>
      <xmax>328</xmax>
      <ymax>344</ymax>
    </bndbox>
  </object>
</annotation>

[Reagan1311]

I found the segmentation annotation in this repository https://github.com/ivalab/affordanceNet_DA. It seems the number of images are different in these two links: dropbox1 dropbox2 (66043 vs 37200). What is the difference?

[acmiyaguchi]

Maybe @ruinianxu would be able to give context?

As far as I know, this repository pairs with https://github.com/ivalab/simData_imgSaver which has some scripts for generating affordances. There might have been some changes between that and the affordanceNet work.