hjkwon0609 / source_separation_ml_jeju Goto Github PK
View Code? Open in Web Editor NEWSource Separation Project For ML Jeju Camp 2017
Source Separation Project For ML Jeju Camp 2017
I debugged the program, and find that the batch_size is 18000(frames).
(18000+3)*256/16000 = 288s, isn't this too long?
According to the idea of paper
'''
Stack spectrograms so that each element in the spectrogram now has 3 elements (prev_frame, curr_frame, next_frame)
For the first(last) frame, prev_frame(next_frame) is zero vector
'''
stacked = tf.stack([tf.pad(spec,[[0, 2], [0, 0]], "CONSTANT"),
tf.pad(spec,[[1, 1], [0, 0]], "CONSTANT"),
tf.pad(spec,[[0, 2], [0, 0]], "CONSTANT")], axis=2) # axis=0
I think the code should be:
stacked = tf.stack([tf.pad(spec,[[2, 0], [0, 0]], "CONSTANT"),
tf.pad(spec,[[1, 1], [0, 0]], "CONSTANT"),
tf.pad(spec,[[0, 2], [0, 0]], "CONSTANT")], axis=2) # axis=0
Am I correct?
hi,
I just want to use the model,
could u pls supply the address ?
thx
hi,
In the preprocessing.py, it's no need to set the sr=44100 at librosa.load and then resample,
U can directly set sr=16000 in librosa.load
default codes:
data, rate = librosa.load(filename, mono=False, sr=44100)
song = data[0,:]
voice = data[1,:]
song = librosa.resample(song, rate, 16000)
voice = librosa.resample(voice, rate, 16000)
after modified:
data, rate = librosa.load(filename, mono=False, sr=16000)
song = data[0,:]
voice = data[1,:]
thx
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.