Robust Vision Challenge 2018

In conjunction with CVPR 2018

Submission Guidelines

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.

Submission Process

We are very much looking forward to your submission. To submit to ROB 2018, follow these steps:

  1. Download the development kit using the link below
  2. Download and convert all datasets into a common format using the devkit
  3. Train your model on the unified training set and run your trained model on the unified test set
  4. Use the devkit to convert your results back to the original dataset format(s)
  5. Choose a short method name (between 4 and 16 characters excluding suffix) for your method (allowed characters: + - _ A..Z a..z 0..9)
  6. Append the suffix _ROB to your method name
  7. Upload your results to each indivdiual benchmark
  8. Register your method to the ROB challenge using the form below
Registering a method triggers our crawler and ranking script, after a few minutes the tables should be updated (see timestamp below the tables). In addition, the script is executed once a day at midnight CET. If you update your already registered results you will therefore see the method re-ranked the next day.

Challenges

ROB 2018 features 6 challenges: stereo, multi-view stereo (MVS), optical flow, single image depth prediction, semantic segmentation and instance segmentation. Participants are free to submit to a single challenge or to multiple challenges. For each challenge, the results of a single model must be submitted to all benchmarks (indicated with an x below). For each challenge we will award a price to the winner and the runner-up.

Stereo
MVS
Flow
Depth
Semantic
Instance
✘ ✘

ROB Development Kit

For downloading the ROB devkit please provide your email, we will send you download instructions. We will use your email address solely for sending you updates about the status of the challenge and the associated workshop. We will not share your email address with any third party or use it for any other purpose.

Register Method to ROB

For downloading the ROB development kit, please provide your email below. We will use your email address solely for sending you updates about the status of the challenge and the associated workshop. We will not share your email address with any third party or use it for any other purpose.

I have read and I accept the submission guidelines.

Fields marked with an asterisk (*) are mandatory.


Sponsors



eXTReMe Tracker