Please carefully read the following guidelines before submitting. If you do not comply with our guidlines your method will be removed.
The goal of ROB is to foster development of algorithms which are robust across various datasets. Thus, each participating method must be trained and tested on all datasets involved in the respective challenge. It is not allowed to use different methods or to alter the parameters of a method for each individual benchmark of a challenge. By submitting to ROB you agree to hand your source code to the organizing team upon request for inspection. Moreover, a paper, technical report or Arxiv paper must be published by the time of the workshop which covers the algorithmic details of the approach. External training data may be used as long as the dataset is public. The winner and the runner-up of each category will receive prize money, are invited to present the method at the ROB 2018 workshop and participate in a joint dinner, as well as co-author a joint TPAMI submission. Participants may submit to a single or multiple challenges.
Invalid Submission Examples:
- A method which trains a classifier to detect from which benchmark a file comes from (either by image content or meta data, e.g., image dimensions)
- A method which executes separate program paths specifically designed for individual datasets (this excludes pre-processing such as image resizing which is allowed)
- A method which is trained on non-publicly available datasets
Valid Submission Examples:
- A method that was trained with all training/validation data available from the individual benchmarks and additional public data (eg, pre-trained on ImageNet or Mapillary Vistas) but does not contain dataset specific instructions or training
- A method which learns to classify the dataset by the image content alone and then solves the task per dataset individually but without providing explicit supervision for classification. In particular, the classifier part may not be trained separately from the other tasks and no meta-information may be used anywhere to identify datasets. The dataset specific network may also not be (pre-)trained on individual datasets alone (no meta-information such as the image dimensions or the dataset an image comes from may be used).
- A method is trained with the ROB training set in a supervised manner and with the ROB test set in an unsupervised manner.