If you submit a stereo matching algorithm, please submit a zip file named <last-name_first-name_method-name>.zip containing the following:
  • A readme.txt containing your full name, email, and organization, along with a short description of your method and any additional information you feel is necessary
  • A prepackaged binary, python script, or MATLAB program that will be run in a VM on an Ubuntu 16.04 LTS Server Edition (64-bit) Machine. Specs:
    • CPU: 2x Intel Xeon E5-2630 2.2GHz
    • Memory: 128GB DDR4-2400
    • GPU: 2x NVIDIA Titan X 12GB
  • Your method must be able to be run by typing...
    • (Binary) "<binary-name> <left-rectified-image> <right-rectified-image>" and output a disparity image
    • (Python) "python <script-name>.py <left-rectified-image> <right-rectified-image>" and output a disparity image
    • (MATLAB) Will be run by calling your main function from the MATLAB interpreter with "<your-function-name>(<left-rectified-image>,<right-rectified-image>)" and output a disparity image
    • NOTE: If you want to perform your own rectification, please indicate that in your submission readme. Everything will still be run the same, except with raw images.
If there are any questions, please contact the CATS team to discuss.
Please submit .zip files to chandrak@udel.edu with the subject line "CATS Submission"