andrewowens / multisensory

Code for the paper: Audio-Visual Scene Analysis with Self-Supervised Multisensory Features

http://andrewowens.com/multisensory/

andrewowens/multisensory Issues

difference between "large" and "full" sep models
Updated 2 years ago
Download sample-data.zip NOT FOUND
Updated 2 years ago1
model architecture
Updated 3 years ago1
Download pretrain models
Updated 3 years ago3
Question about the original audio waveform input
Closed 3 years ago
How to calculate SDR?
Updated 4 years ago1
Question about the test in Table 2 GRID transfer
Updated 4 years ago
question about sourcesep training result on new dataset
Updated 4 years ago7
Why acc doesn't change when shift_model training?
Updated 4 years ago10
Questions about sourcesep.py
Updated 4 years ago2
About the input format
Updated 4 years ago6
file input for blind audio source separation
Updated 4 years ago1
Improvement on using pretrained model
Updated 4 years ago
Questions about VoxCeleb2 dataset
Updated 4 years ago1
What GPU used?
Closed 4 years ago
Issue with sound source localization
Updated 4 years ago1
duration_mult flag
Updated 5 years ago
Issue on datasets
Updated 5 years ago1
Issue on Large Videos
Closed 5 years ago2
Some questions about training and testing shift model
Updated 5 years ago
Question about the test in Table 3
Updated 5 years ago1
What are feats['im_0'] and feats['im_1'] of example for shift model?
Closed 5 years ago2
question about using 'sep_example.tf'
Updated 5 years ago2
Could you provide the dataset?
Closed 5 years ago
About the input file for shift model training
Updated 5 years ago
Questions about the files in ".txt" format used to train the "shift" model
Updated 5 years ago2
Question about the shift_net.py' training
Updated 5 years ago3
In which way the video frames combine
Closed 5 years ago10
What is the format of the tensor in the code?
Closed 5 years ago
I RuntimeError: Command failed! ffmpeg -i "/tmp/ao_wmjz0ezg.wav" -r 29.970000 -loglevel warning -safe 0 -f concat -i "/tmp/ao_i2pwi0b8.txt" -pix_fmt yuv420p -vcodec h264 -strict -2 -y -acodec aac "results/fg_translator.mp4"
Closed 5 years ago4
RuntimeError: Command failed! ffmpeg -i "/tmp/ao_M0QAze.wav" -r 29.970000 -loglevel warning -safe 0 -f concat -i "/tmp/ao_cnpblR.txt" -pix_fmt yuv420p -vcodec h264 -strict -2 -y -acodec aac "../results/fg_cam_translator.mp4"
Closed 6 years ago4
whre is the sep_module (calss or funtion）in sourcesep.py
Updated 5 years ago3
Question about training
Updated 5 years ago10
> > In the source separation model it seems like you are using *.tf files as input (rec_files_from_path in sep_dset.py).Can you please provide the format to create those TFRecord files
Updated 5 years ago
How to train the "shift" and "cam" model for sound source location?
Updated 5 years ago4
Questions about the entrance of the training function
Updated 5 years ago1
TypeError: convolution() got multiple values for argument 'weights_regularizer'
Closed 5 years ago5
Test set used in paper
Updated 6 years ago
error compiling
Closed 6 years ago
make_video_helper() missing 3 required positional arguments: 'x', 'in_dir', and 'tmp_ext'
Updated 6 years ago1
Questions about the models
Updated 6 years ago3
Question about fine-tune for full sep model
Updated 6 years ago3
Supported on Linux
Updated 6 years ago
How do I run source separation on a different video?
Closed 6 years ago2
Getting /bin/sh: 1: ffmpeg-length: not found
Closed 6 years ago2