issues
search
YuanGongND
/
ast
Code for the Interspeech 2021 paper "AST: Audio Spectrogram Transformer".
BSD 3-Clause "New" or "Revised" License
1.17k
stars
221
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
How to configure the dataset or modify the code if I want to do the one class binary classification
#91
nanyyyyyy
opened
1 year ago
14
Can AST be used for audio representation towards solving the frame-level classification tasks?
#90
SylviaZiyaZhou
opened
1 year ago
4
Cannot create a file when that file already exists: './exp/test-esc50-f10-t10-impTrue-aspTrue-b48-lr1e-5/fold1/models'
#89
JonathanFL
closed
1 year ago
4
About code"100-106" from dataloader.py
#88
poult-lab
opened
1 year ago
6
Audio length 1s
#87
9B8DY6
closed
2 years ago
5
Q: Audio-MAE reports that AST's performance on VoxCeleb1 is 41.1%, but not listed on the AST paper
#86
daisukelab
closed
2 years ago
2
The problem of reproducing the AST result in full dataset
#85
MichaelLynn1996
opened
2 years ago
9
Input type (torch.FloatTensor) and weight type (torch.cuda.HalfTensor) should be the same or input should be a MKLDNN tensor and weight is a dense tensor
#84
michelle-chou25
opened
2 years ago
15
validate_ensemble(args, epoch):
#83
shasso2s
opened
2 years ago
1
Sorry, I have a stupid question about how to download audioset
#82
liyunlongaaa
closed
1 year ago
7
Some questions about the details of AST.
#81
TungyuYoung
opened
2 years ago
1
'ASTModel' object has no attribute 'module'
#80
couragelfyang
opened
2 years ago
6
About random noise which author put on the speech command.
#79
poult-lab
closed
2 years ago
11
Unknown model (vit_deit_base_distilled_patch16_384)
#78
BaronWang0130
opened
2 years ago
3
some question about Deit's two [cls] token processing.
#77
liyunlongaaa
opened
2 years ago
2
Using mel spectrogram with different number of bins
#76
Yuhan-Shen
closed
2 years ago
2
The validation loss seems too high
#75
haoheliu
closed
2 years ago
4
AST tiny and small pretrained models
#74
abaronetto
opened
2 years ago
1
Training ESC50 on constrained GPU resources
#73
adrianSRoman
closed
2 years ago
8
normalization
#72
caizihui
closed
2 years ago
0
why not mean of class-wise accuracy ?
#71
liyunlongaaa
closed
2 years ago
2
How do I resume training after an unexpected interruption in training?
#70
liyunlongaaa
closed
2 years ago
2
require torchvision 0.9.1
#69
almostimplemented
closed
2 years ago
0
[Bug] ModuleNotFoundError: No module named 'torch.ao'
#68
almostimplemented
opened
2 years ago
7
[Question] Question about padding operation
#67
Mountchicken
opened
2 years ago
3
Run ESC-50 fine tuned model on test data with metrics
#66
p4vlos
closed
2 years ago
5
Different train-(val/test) spectogram shape (recordings duration)
#65
danihinjos
opened
2 years ago
8
support distribution training with multi GPU ?
#64
joewale
opened
2 years ago
0
Test AST on eeg data
#63
Hitesh-Kumar-2001
opened
2 years ago
2
Overfitting? test AST on GTZAN
#62
kelvinqin
opened
2 years ago
23
Question Regarding Activation on MLP Head
#61
arshinmar
closed
2 years ago
1
Poor performance of AST on audio clips with different lengths
#60
Yuanbo2020
opened
2 years ago
3
Attention maps for model explainability
#59
kremHabashy
closed
2 years ago
1
About start training: IndexError: tuple index out of range.
#58
TungyuYoung
opened
2 years ago
13
Does the input necessarily need to be normalized according to a certain mean and variance?
#57
Basums
closed
2 years ago
3
Question about json file and label index
#56
TungyuYoung
closed
2 years ago
8
Question normalize operation
#55
liuyoude
closed
2 years ago
2
OSError: ./exp/test-esc50-f10-t10-impTrue-aspTrue-b48-lr1e-5/fold1/result.csv not found.(ESC-50 Recipe)
#54
poult-lab
closed
2 years ago
8
How to use the model for a downstream task ?
#53
devesh-k
closed
2 years ago
9
Question regarding fbank for fine tuning
#52
kremHabashy
closed
2 years ago
5
Wonderful work! questions about feature size
#51
lijuncheng16
closed
2 years ago
7
Convert mel filterbanks to wav again?
#50
clairerity
closed
2 years ago
1
Prediction always wrong using esc50 recipe with 0.95+ accuracy after training
#49
kremHabashy
closed
2 years ago
2
Normalizing the train and test data
#48
ranjith1604
closed
2 years ago
3
Question about pre-training on a new dataset.
#47
devesh-k
closed
2 years ago
5
Temporal organization of tokens
#46
MikeKras
closed
2 years ago
2
Positional embedding
#45
lmaxwell
closed
2 years ago
2
How to change the interpolation method?
#44
ooobsidian
closed
2 years ago
6
How to change the kernel size?
#43
ooobsidian
closed
2 years ago
2
results.csv and getting labels per audio file
#42
Gsantos4
closed
2 years ago
1
Previous
Next