qiuqiangkong/audioset_tagging_cnn

Issues

• Share audioset/dataset on Huggingface?

  #73 opened 3 months ago by IvanBirkmaier
  0

• How can i download the DataSet ?

  #23 opened 4 years ago by jiguanglu
  13

• The download address of the pretrained model cannot be found

  #71 opened 5 months ago by kuikui12385128149
  0

• numba.decorators ModuleNotFoundError

  #46 opened 3 years ago by onnkeat
  3

• "/metadata/unbalanced_partial_csvs/unbalanced_train_segments_part$IDX.csv" does not exists.

  #62 opened a year ago by Dxymiemiemie
  1

• Unable to decompress AudioSet on BaiduPan

  #70 opened 9 months ago by TFEI-Nagato
  0

• Where are DecisionLevel explained?

  #69 opened 10 months ago by JNaranjo-Alcazar
  0

• is it possible to turn clipwise model into framewise w/o retrain?

  #67 opened a year ago by chamecall
  2

• How to deal with "File does not exist"?

  #68 opened 10 months ago by maxwZJU
  0

• DecisionLevelMax with mobilenet2

  #24 opened 4 years ago by unrelatedlabs
  1

• About Model Saving

  #66 opened a year ago by AntyRia
  1

• Realtime detection of sound

  #65 opened a year ago by giusarno
  0

• how to fine-tune an existing model with my own data set?

  #64 opened a year ago by moseranliunian
  0

• Difficulty Downloading AudioSet from Pan Baidu Storage for Non-Chinese Users

  #63 opened a year ago by gevmin94
  1

• inconsistent result in batch mode

  #61 opened 2 years ago by manish-kumar-iisc
  0

• about some use question

  #60 opened 2 years ago by AntyRia
  1

• Get embedding not classification

  #51 opened 3 years ago by newusersayshi
  1

• Shape doesn't match when inferencing Wavegram_Logmel_Cnn14 model

  #59 opened 2 years ago by LllHan
  1

• wav file not include in eval_segment.csv

  #57 opened 2 years ago by zdj97
  4

• Cnn14_DecisionLevelMax for 16kHz audio

  #58 opened 2 years ago by kremnik
  0

• Factor of 2 in count_flops.py

  #56 opened 2 years ago by fschmid56
  0

• Feeding long audio data vs second-by-second or smaller chunks

  #50 opened 3 years ago by anarsultani97
  2

• Error loading the pre-trained model

  #55 opened 2 years ago by outman-goutian
  1

• how about use Cnn14_8k_mAP=0.438.pth

  #54 opened 3 years ago by herbiel
  0

• panns_transfer_to_gtzan数据集链接失效

  #53 opened 3 years ago by wusuowei11
  0

• Last dropout is disconnected from fc_audioset layer

  #52 opened 3 years ago by ilic-mezza
  0

• Can use 32k mobilenetv2 model to fineturn 16k mobilenetv2 model?

  #49 opened 3 years ago by AricJames
  0

• Can not reproducing the result of audio_tagging result of mobilenetv1 in the PANNs paper, is there any tricks when training?

  #48 opened 3 years ago by ccjluff
  0

• batchnorm1d doesn't seem to be used in attention block

  #47 opened 3 years ago by filipmu
  0

• have u ever tried transformer based model

  #45 opened 3 years ago by chichixdf
  0

• Coud not find file "unbalanced_train_segments_part38_partial.z01" in dataset

  #44 opened 3 years ago by owlwang
  2

• Variable Length Sequences

  #43 opened 3 years ago by zhaowil
  1

• Binarizing output values

  #41 opened 3 years ago by anarsultani97
  1

• object of type 'NoneType' has no len()

  #38 opened 4 years ago by nkw97
  1

• [asdf problem] Hello, is this typo in "utils/plot_statistics.py " ?

  #42 opened 3 years ago by yeopee2
  0

• Literature pointers for better understanding the `Cnn14_DecisionLevelAtt` model

  #32 opened 4 years ago by sirgarfieldc
  11

• Is it possible to put the spec_augmenter in front of the nn.BatchNorm2d

  #40 opened 4 years ago by garrywrj
  4

• Why transpose(1, 3) before BatchNorm?

  #39 opened 4 years ago by ado-ml
  1

• convert the sound detection event predicting image into csv (Pandas format)

  #35 opened 4 years ago by SR-fr
  1

• code to plot "log spectrogram"+class probabilities

  #33 opened 4 years ago by gewueh
  1

• The procedure

  #30 opened 4 years ago by IcyHao
  24

• First float32_to_int16, and then int16_to_float32?

  #37 opened 4 years ago by sysuzyq
  2

• The sample rate of CNN14_emb128_mAP=0.412.pth

  #34 opened 4 years ago by Usernamezhx
  1

• Input wav's time length for model "Wavegram_Logmel_Cnn14"?

  #29 opened 4 years ago by machengnan
  1

• ERROR - code is too big

  #27 opened 4 years ago by shiyu22
  1

• ERROR: -0BIyqJj9ZU: YouTube said: Invalid parameters.

  #25 opened 4 years ago by 2017csz0006
  1

• The procedure

  #31 opened 4 years ago by IcyHao
  0

• Other Pretrained Models

  #28 opened 4 years ago by PhenomenalOnee
  1

• Could i change the input size of wav files?

  #26 opened 4 years ago by xinmokunshou
  0

• What's the input size of CNN

  #21 opened 4 years ago by xiaoluobu
  2