Participants are invited to submit results and a short description of the used method(s) together with a signed form. We encourage participants to provide results based on stand-alone video data, stand-alone device data and a You of the two, though is not required and participants are free to just provide results for one or two the three possibilities. Participants are free to use third party public data to augmented the provided training data. Participants are encouraged, but not required, to also submit results without pretraining to allow future comparisons. We only allow submissions from methods that do not rely on information from future frames to predict the current phase, i.e. online workflow analysis.
Please follow the instructions in the "Readme" file provided with the test data.
Please upload your results for the phase detection, the signed honor code and the description of the method used in a zip archive. Please indicate whether video, device or combined data were used to compute the results. <Username>_Results_[Video|Device|Combined].zip.
Please provide a csv file (Filename: <OriginalName>_res.csv) for each operation in the following format:
<frame number>,<detected phase id> e.g.: 0,0 1,0 ... 2453,5 2454,5 ...
(Note: We do not require you to submit results on instrument usage, nor do we evaluate this)
Please upload your submission here.
Note that you can only upload if you are registered as a user for the sub-challenge. If you have problems uploading, please send the zip file to sebastian.bodenstedt (at) nct-dresden.de.