******************************************************************
      Instruction for REVERB Challenge SE result submission 
  
                                             REVERB challenge organizers
					  reverb-challenge@lab.ntt.co.jp
      ******************************************************************

- Please first select "number of microphones" and
  "type of processing scheme (i.e., full batch processing,
  utterance-based batch processing, real-time processing)".
  of your algorithm using the pull-down menu on the left side
  of the tables.

- We prepared one excel sheet for each objective measure.
  Since cepstrum distance, log likelihood ratio (LLR),
  frequency-weighted segmental SNR (FWSegSNR), and
  speech-to-reverberation modulation energy ratio (SRMR) are
  the mandatory evaluation criteria, you are asked to fill out
  the tables corresponding to these criteria.

- If you would like to submit results for various number of 
  microphones and processing schemes (i.e., full batch processing, 
  utterance-based batch processing, real-time processing), please add
  additional row(s) to the table to describe the results.
  Note that you can submit only 1 result per number of microphones and 
  processing scheme.
  You can include additional results in your workshop paper.

- We strongly encourage participants to mention computational
  complexity of employed algorithm on a sheet named
  "wall clock time (optional)". It is recommended to report
  computational time consumed to process the data (i.e., wall clock 
  time). For the details about the calculation of wall clock time,
  please see http://reverb2014.dereverberation.com/download.html.
  If you wish to report computational complexity
  in a different manner, please describe it in the same excel
  sheet in your own way.

- If you have PESQ score, please fill out the table on
  the "PESQ (optional)" sheet.

- If you have ASR score (word error rate), please use the ASR
  result submission form. For more details, please refer to the readme
  file for ASR result submission format.

- As it is mentioned on the webpage, we are going to perform
  subjective evaluation of the audio data processed by participants.
  Regarding the audio data submission, we will contact
  you again and tell which audio data to submit, and how and when
  to submit them.
  Note that you must submit the audio data that is used to
  calculate the final objective SE scores. It is not allowed to
  tune your algorithms for subjective evaluation separately from
  the objective evaluation.

- When you submit the result to us, please let us know your contact 
  information (name and email) and affiliation.