Juntae, Kim

Good news! a full python implementation is upcoming (to be updated until 2018/03/17)

Sorry for the late update, we already have a fully python written test script however, the train script has been delayed. We will firstly update test script in this week...

* NOTE, we uploaded the test script fully written by python in py branch.

Unfortunately, there is no automatic memory management function in this project. Therefore, you should manually split the long wave file into some several short wave file of which the size...

Actually the training set of our vad doesn't have music sound. The result you uploaded can become better if some post-processing is applied. But we have only matlab version post...

@vladfulgeanu We upload the end-point detection (EPD) algorithm to https://github.com/jtkim-kaist/end-point-detection (it may be you want because EPD find the start and end point of the speech signal.) However, the used...

Actually we are currently researching the small size VAD for embedded system (CNN based). The current VAD in this github repos. is not proper for embedded system. Anyway we have...

Thank you for your detailed question. My advice is, In VAD field, I experimentally found that dropout seems not to be necessary as we use already noise corrupted data, which...

Thank you for your sharing. My question is 1. What is your validation set? (Also, the higher accuracy of validation set is not the problem as it can be changed...

Thank you for sharing. Note that there is some additional tricks apart from the training method described in my paper (because of paper limitation). Training noisy data is hard problem...