[Company] [Ivory Tower] PPG Ali paper version. The tacotron part, draw the localization algorithm. Draw the form to design the hyperparameter search experiment. Discuss using rayhane or torch. PPG extraction and write the algorithm, by the way, train the cpu version, and learn Source PPT. Enumeration of PPT schemes completed before 1/31
[Ivory Tower] Mixed language stage2, LID correction, pre-processing as a topic . Then consider the mixed text processing plan (or not processing first), see the note
[Company] [Ivory Tower] AutoVC (sil, dataloader, val cutting), see the results, re-adjust parameters, one-hot version, hyperparameter comparison. Add the output during training to wav, val and conversion. No wavenet is used as output , Write GL. Use the neural network to do it again, add a new version, improve the blog, analyze, and give the experience and skills of AutoVC~. Deadline at 5:00 pm on Monday, 1/18, give Demo
[School] [ivory tower] Show the PPT of the accented paper and read the paper again, plus the parallel corpus of the data set. Go to the laboratory to find the source at 8:00 am on 1/19
[School] Summarize the PPG Project method as a patent. The content of the PPT is organized into a blog to sort out some things. Discuss with Xueyuan teacher before 1/22
[Old School] Organize Xintao's papers
[Old School] Find the latest paper recommended by the teacher to read and organize
[Company] Maybe-TTS is transferred according to the module, the ASR result and the py file of the merged text to calculate the pause , and then continue
[Ivory Tower] [Company] Meet Yelaoshr’s Melgan
Today's task
[Company] [Ivory Tower] AutoVC (sil, dataloader, val cutting), one-hot version, hyperparameter comparison. Added the output during training to wav, val and conversion. GL finds bugs. The version with the sound from colleagues For benchmark reproduction, give DEMO
[School] [ivory tower], the data set parallel corpus display, stress the PPT and the paper read again. Asked Brother Jingbei
[Company] Maybe-TTS, ASR results and the py file that merges the pause text. That's it!