Code for the Interspeech 2021 paper "AST: Audio Spectrogram Transformer".

Related tags

Deep Learningast
Overview

AST: Audio Spectrogram Transformer

Introduction

Illustration of AST.

This repository contains the official implementation (in PyTorch) of the Audio Spectrogram Transformer (AST) proposed in the Interspeech 2021 paper AST: Audio Spectrogram Transformer (Yuan Gong, Yu-An Chung, James Glass).

AST is the first convolution-free, purely attention-based model for audio classification which supports variable length input and can be applied to various tasks. We evaluate AST on various audio classification benchmarks, where it achieves new state-of-the-art results of 0.485 mAP on AudioSet, 95.6% accuracy on ESC-50, and 98.1% accuracy on Speech Commands V2. For details, please refer to the paper and the ISCA SIGML talk.

Please have a try! AST can be used with a few lines of code, and we also provide recipes to reproduce the SOTA results on AudioSet, ESC-50, and Speechcommands with almost one click.

The AST model file is in src/models/ast_models.py, the recipes are in egs/[audioset,esc50,speechcommands]/run.sh, when you run run.sh, it will call /src/run.py, which will then call /src/dataloader.py and /src/traintest.py, which will then call /src/models/ast_models.py.

Citing

Please cite our paper(s) if you find this repository useful. The first paper proposes the Audio Spectrogram Transformer while the second paper describes the training pipeline that we applied on AST to achieve the new state-of-the-art on AudioSet.

@article{gong2021ast,  
 title={Ast: Audio spectrogram transformer}, 
 author={Gong, Yuan and Chung, Yu-An and Glass, James}, 
 journal={arXiv preprint arXiv:2104.01778}, 
 year={2021}}  
@article{gong2021psla,  
 title={PSLA: Improving Audio Tagging with Pretraining, Sampling, Labeling, and Aggregation}, 
 author={Gong, Yuan and Chung, Yu-An and Glass, James}, 
 journal={arXiv preprint arXiv:2102.01243}, 
 year={2021}}  

Getting Started

Step 1. Clone or download this repository and set it as the working directory, create a virtual environment and install the dependencies.

cd ast/ 
python3 -m venv venvast
source venvast/bin/activate
pip install -r requirements.txt 

Step 2. Test the AST model.

ASTModel(label_dim=527, \
         fstride=10, tstride=10, \
         input_fdim=128, input_tdim=1024, \
         imagenet_pretrain=True, audioset_pretrain=False, \
         model_size='base384')

Parameters:
label_dim : The number of classes (default:527).
fstride: The stride of patch spliting on the frequency dimension, for 16*16 patchs, fstride=16 means no overlap, fstride=10 means overlap of 6 (used in the paper). (default:10)
tstride: The stride of patch spliting on the time dimension, for 16*16 patchs, tstride=16 means no overlap, tstride=10 means overlap of 6 (used in the paper). (default:10)
input_fdim: The number of frequency bins of the input spectrogram. (default:128)
input_tdim: The number of time frames of the input spectrogram. (default:1024, i.e., 10.24s)
imagenet_pretrain: If True, use ImageNet pretrained model. (default: True, we recommend to set it as True for all tasks.)
audioset_pretrain: IfTrue, use full AudioSet And ImageNet pretrained model. Currently only support base384 model with fstride=tstride=10. (default: False, we recommend to set it as True for all tasks except AudioSet.)
model_size: The model size of AST, should be in [tiny224, small224, base224, base384] (default: base384).

cd ast/src
python
import os 
import torch
from models import ASTModel 
# download pretrained model in this directory
os.environ['TORCH_HOME'] = '../pretrained_models'  
# assume each input spectrogram has 100 time frames
input_tdim = 100
# assume the task has 527 classes
label_dim = 527
# create a pseudo input: a batch of 10 spectrogram, each with 100 time frames and 128 frequency bins 
test_input = torch.rand([10, input_tdim, 128]) 
# create an AST model
ast_mdl = ASTModel(label_dim=label_dim, input_tdim=input_tdim, imagenet_pretrain=True)
test_output = ast_mdl(test_input) 
# output should be in shape [10, 527], i.e., 10 samples, each with prediction of 527 classes. 
print(test_output.shape)  

ESC-50 Recipe

The ESC-50 recipe is in ast/egs/esc50/run_esc.sh, the script will automatically download the ESC-50 dataset and resample it to 16kHz, then run standard 5-cross validation and report the result. The recipe was tested on 4 GTX TITAN GPUs with 12GB memory. The result is saved in ast/egs/esc50/exp/yourexpname/acc_fold.csv (the accuracy of fold 1-5 and the averaged accuracy), you can also check details in result.csv and best_result.csv (accuracy, AUC, loss, etc of each epoch / best epoch). We attached our log file in ast/egs/esc50/test-esc50-f10-t10-p-b48-lr1e-5, the model achieves 95.75% accuracy.

To run the recipe, simply comment out . /data/sls/scratch/share-201907/slstoolchainrc in ast/egs/esc50/run_esc.sh, adjust the path if needed, and run:

cd ast/egs/esc50
(slurm user) sbatch run_esc50.sh
(local user) ./run_esc50.sh

Speechcommands V2 Recipe

The Speechcommands recipe is in ast/egs/speechcommands/run_sc.sh, the script will automatically download the Speechcommands V2 dataset, train an AST model on the training set, validate it on the validation set, and evaluate it on the test set. The recipe was tested on 4 GTX TITAN GPUs with 12GB memory. The result is saved in ast/egs/speechcommands/exp/yourexpname/eval_result.csv in format [val_acc, val_AUC, eval_acc, eval_AUC], you can also check details in result.csv (accuracy, AUC, loss, etc of each epoch). We attached our log file in ast/egs/speechcommends/test-speechcommands-f10-t10-p-b128-lr2.5e-4-0.5-false, the model achieves 98.12% accuracy.

To run the recipe, simply comment out . /data/sls/scratch/share-201907/slstoolchainrc in ast/egs/esc50/run_sc.sh, adjust the path if needed, and run:

cd ast/egs/speechcommands
(slurm user) sbatch run_sc.sh
(local user) ./run_sc.sh

Audioset Recipe

Audioset is a little bit more complex, you will need to prepare your data json files (i.e., train_data.json and eval_data.json) by your self. The reason is that the raw wavefiles of Audioset is not released and you need to download them by yourself. We have put a sample json file in ast/egs/audioset/data/datafiles, please generate files in the same format (You can also refer to ast/egs/esc50/prep_esc50.py and ast/egs/speechcommands/prep_sc.py.). Please keep the label code consistent with ast/egs/audioset/data/class_labels_indices.csv.

Once you have the json files, you will need to generate the sampling weight file of your training data (please check our PSLA paper to see why it is needed).

cd ast/egs/audioset
python gen_weight_file.py ./data/datafiles/train_data.json

Then you just need to change the tr_data and te_data in /ast/egs/audioset/run.sh and then

cd ast/egs/audioset
(slurm user) sbatch run.sh
(local user) ./run.sh

You should get a model achieves 0.448 mAP (without weight averaging) and 0.459 (with weight averaging). This is the best single model reported in the paper. The result of each epoch is saved in ast/egs/audioset/exp/yourexpname/result.csv in format [mAP, mAUC, precision, recall, d_prime, train_loss, valid_loss, cum_mAP, cum_mAUC, lr] , where cum_ results are the checkpoint ensemble results (i.e., averaging the prediction of checkpoint models of each epoch, please check our PSLA paper for details). The result of weighted averaged model is saved in wa_result.csv in format [mAP, AUC, precision, recall, d-prime]. We attached our log file in ast/egs/audioset/test-full-f10-t10-pTrue-b12-lr1e-5/, the model achieves 0.459 mAP.

In order to reproduce ensembe results of 0.475 mAP and 0.485 mAP, please train 3 models use the same setting (i.e., repeat above three times) and train 6 models with different tstride and fstride, and average the output of the models. Please refer to ast/egs/audioset/ensemble.py. We attached our ensemble log in /ast/egs/audioset/exp/ensemble-s.log and ensemble-m.log. You can use our pretrained models (see below) to test ensemble result.

Pretrained Models

We provide full AudioSet pretrained models.

  1. Full AudioSet, 10 tstride, 10 fstride, with Weight Averaging (0.459 mAP)
  2. Full AudioSet, 10 tstride, 10 fstride, without Weight Averaging, Model 1 (0.450 mAP)
  3. Full AudioSet, 10 tstride, 10 fstride, without Weight Averaging, Model 2 (0.448 mAP)
  4. Full AudioSet, 10 tstride, 10 fstride, without Weight Averaging, Model 3 (0.448 mAP)
  5. Full AudioSet, 12 tstride, 12 fstride, without Weight Averaging, Model (0.447 mAP)
  6. Full AudioSet, 14 tstride, 14 fstride, without Weight Averaging, Model (0.443 mAP)
  7. Full AudioSet, 16 tstride, 16 fstride, without Weight Averaging, Model (0.442 mAP)

Ensemble model 2-4 achieves 0.475 mAP, Ensemble model 2-7 achieves and 0.485 mAP. You can download these models at one click using ast/egs/audioset/download_models.sh. Once you download the model, you can try ast/egs/audioset/ensemble.py, you need to change the eval_data_path and mdl_list to run it. We attached our ensemble log in /ast/egs/audioset/exp/ensemble-s.log and ensemble-m.log.

If you want to finetune AudioSet-pretrained AST model on your task, you can simply set the audioset_pretrain=True when you create the AST model, it will automatically download model 1 (0.459 mAP). In our ESC-50 recipe, AudioSet pretraining is used.

Contact

If you have a question, please bring up an issue (preferred) or send me an email [email protected].

Owner
Yuan Gong
Ph.D in CS
Yuan Gong
Unofficial reimplementation of ECAPA-TDNN for speaker recognition (EER=0.86 for Vox1_O when train only in Vox2)

Introduction This repository contains my unofficial reimplementation of the standard ECAPA-TDNN, which is the speaker recognition in VoxCeleb2 dataset

Tao Ruijie 277 Dec 31, 2022
Semantic graph parser based on Categorial grammars

Lambekseq "Everyone who failed Greek or Latin hates it." This package is for proving theorems in Categorial grammars (CG) and constructing semantic gr

10 Aug 19, 2022
OrienMask: Real-time Instance Segmentation with Discriminative Orientation Maps

OrienMask This repository implements the framework OrienMask for real-time instance segmentation. It achieves 34.8 mask AP on COCO test-dev at the spe

45 Dec 13, 2022
An example of Scatterbrain implementation (combining local attention and Performer)

An example of Scatterbrain implementation (combining local attention and Performer)

HazyResearch 97 Jan 02, 2023
Code for the paper "Adapting Monolingual Models: Data can be Scarce when Language Similarity is High"

Wietse de Vries • Martijn Bartelds • Malvina Nissim • Martijn Wieling Adapting Monolingual Models: Data can be Scarce when Language Similarity is High

Wietse de Vries 5 Aug 02, 2021
Code for "Diffusion is All You Need for Learning on Surfaces"

Source code for "Diffusion is All You Need for Learning on Surfaces", by Nicholas Sharp Souhaib Attaiki Keenan Crane Maks Ovsjanikov NOTE: the linked

Nick Sharp 247 Dec 28, 2022
Code for Multiple Instance Active Learning for Object Detection, CVPR 2021

Language: 简体中文 | English Introduction This is the code for Multiple Instance Active Learning for Object Detection, CVPR 2021. Installation A Linux pla

Tianning Yuan 269 Dec 21, 2022
Public implementation of "Learning from Suboptimal Demonstration via Self-Supervised Reward Regression" from CoRL'21

Self-Supervised Reward Regression (SSRR) Codebase for CoRL 2021 paper "Learning from Suboptimal Demonstration via Self-Supervised Reward Regression "

19 Dec 12, 2022
Interactive web apps created using geemap and streamlit

geemap-apps Introduction This repo demostrates how to build a multi-page Earth Engine App using streamlit and geemap. You can deploy the app on variou

Qiusheng Wu 27 Dec 23, 2022
Share a benchmark that can easily apply reinforcement learning in Job-shop-scheduling

Gymjsp Gymjsp is an open source Python library, which uses the OpenAI Gym interface for easily instantiating and interacting with RL environments, and

134 Dec 08, 2022
CSPML (crystal structure prediction with machine learning-based element substitution)

CSPML (crystal structure prediction with machine learning-based element substitution) CSPML is a unique methodology for the crystal structure predicti

8 Dec 20, 2022
Towards Implicit Text-Guided 3D Shape Generation (CVPR2022)

Towards Implicit Text-Guided 3D Shape Generation Towards Implicit Text-Guided 3D Shape Generation (CVPR2022) Code for the paper [Towards Implicit Text

55 Dec 16, 2022
rastrainer is a QGIS plugin to training remote sensing semantic segmentation model based on PaddlePaddle.

rastrainer rastrainer is a QGIS plugin to training remote sensing semantic segmentation model based on PaddlePaddle. UI TODO Init UI. Add Block. Add l

deepbands 5 Mar 04, 2022
This is a simple face recognition mini project that was completed by a team of 3 members in 1 week's time

PeekingDuckling 1. Description This is an implementation of facial identification algorithm to detect and identify the faces of the 3 team members Cla

Eric Kwok 2 Jan 25, 2022
PyTorch Implementation of ECCV 2020 Spotlight TuiGAN: Learning Versatile Image-to-Image Translation with Two Unpaired Images

TuiGAN-PyTorch Official PyTorch Implementation of "TuiGAN: Learning Versatile Image-to-Image Translation with Two Unpaired Images" (ECCV 2020 Spotligh

181 Dec 09, 2022
Live training loss plot in Jupyter Notebook for Keras, PyTorch and others

livelossplot Don't train deep learning models blindfolded! Be impatient and look at each epoch of your training! (RECENT CHANGES, EXAMPLES IN COLAB, A

Piotr Migdał 1.2k Jan 08, 2023
Source code for Adaptively Calibrated Critic Estimates for Deep Reinforcement Learning

Adaptively Calibrated Critic Estimates for Deep Reinforcement Learning Official implementation of ACC, described in the paper "Adaptively Calibrated C

3 Sep 16, 2022
RLDS stands for Reinforcement Learning Datasets

RLDS RLDS stands for Reinforcement Learning Datasets and it is an ecosystem of tools to store, retrieve and manipulate episodic data in the context of

Google Research 135 Jan 01, 2023
Google AI Open Images - Object Detection Track: Open Solution

Google AI Open Images - Object Detection Track: Open Solution This is an open solution to the Google AI Open Images - Object Detection Track 😃 More c

minerva.ml 46 Jun 22, 2022
code for our ECCV 2020 paper "A Balanced and Uncertainty-aware Approach for Partial Domain Adaptation"

Code for our ECCV (2020) paper A Balanced and Uncertainty-aware Approach for Partial Domain Adaptation. Prerequisites: python == 3.6.8 pytorch ==1.1.0

32 Nov 27, 2022