Closed dingusagar closed 6 years ago
actually this problem occurs only for the Noise category. When I bypassed the Noise category, feature extraction ran successfully for all other categories - Crying baby , Silence and Baby laugh. The final numpy file was also created after that.
That's weird.
I thought about file type in Noise folder, but it can't be the reason as it is the same kind of file you find in Crying baby folder (.ogg). Moreover, from the log file, all the previous signals take reasonable time to be processed. Is it possible that only one file (5-156999-C.ogg) is corrupted somehow ?
You were right. Some 5 files in the Crying baby folder was corrupted. I removed them from the folder and ran it again. It worked. It generated the output numpy file also. Thanks
train_set.py is running forever.
The log file generated is attached here.
There are 108 files in the directory ./data/902 - Noise. But according to the log file only 53 are getting processed. It gets hung on reading the 53rd ogg file.
2018-02-07 07:07:05 AM - INFO - Reading file: /home/dingu/PycharmProjects/baby_cry_detection/pc_main/../data/902 - Noise/5-156999-C.ogg ...
logs_pc_methods_feat_eng.log