We present the differences between different detector in our paper and analyze the effect of object boxes on HOI detection. VCL detector and DRG detector can be download from the corresponding paper. Due to the space limitation of Google Drive, there are many files provided in CloudStor. Many thanks to CloudStor and The University of Sydney. Here, we provide the GT boxes.
GT boxes annotation: https://drive.google.com/file/d/15UXbsoverISJ9wNO-84uI4kQEbRjyRa8/view?usp=sharing
FCL was finished about 10 months ago. In the first submission, we compare the difference among COCO detector, Fine-tuned Detector and GT boxes. We further find DRG object detector largely increases the baseline. All these comparisons illustrate the significant effect of object detector on HOI. That's really necessary to provide the performance of object detector.
HOI-COCO training data: https://cloudstor.aarnet.edu.au/plus/s/6NzReMWHblQVpht
Please notice train2017 might contain part of V-COCO test data. Thus, we just use train2014 in our experiment. If we use train2017, the result might be better (improve about 0.5%). We think that is the case: we have localized objects, but we do not know the interaction.
Evalation of Object365_COCO: https://cloudstor.aarnet.edu.au/plus/s/GpNkjOHaS8xN5ar
Evalation of COCO (val2017): https://cloudstor.aarnet.edu.au/plus/s/QBXOYiJ5NHqtcti
Evalation of Object365 (novel classes): https://cloudstor.aarnet.edu.au/plus/s/pRBLkhy9TUm5xGo
Evalation of HICO (test): https://cloudstor.aarnet.edu.au/plus/s/AFrv822lPC30iHt
Here we provide the object datasets that we use in this repo. The format is the same as the training data of HICO.
COCO: https://cloudstor.aarnet.edu.au/plus/s/9cgMYKq5B4waawA
Object365_COCO: https://cloudstor.aarnet.edu.au/plus/s/VuHvDdp5msRnpqn. (we do not run this experiment).
COCO (2014 training set): https://cloudstor.aarnet.edu.au/plus/s/N35ovTXWtLmG9ZN
HICO (training set): https://cloudstor.aarnet.edu.au/plus/s/MglGuWmtDgUgJLl
HOI-COCO: https://cloudstor.aarnet.edu.au/plus/s/YEiPiX0B3jaFasU
We also provide all files here: https://cloudstor.aarnet.edu.au/plus/s/9YxboaI6d5lITLk. If I omit some files, you can find the corresponding files in this directory. Noticeably, this directory might include some useless files and broken files that I never use.
I use Cyberduck upload the files to Cloudstor. You might use Cyberduck similar to this (I do not how to download the public files with duck):
duck -u [email protected] --parallel 8 -e compare --download davs://cloudstor.aarnet.edu.au/plus/remote.php/webdav/test/ /tmp/testdata/